One on-campus community I’m involved in is the freshman orientation program. Orientation Leaders serve as mentors for the incoming students over the summer before their freshman year and the first week of their Cornell life. To be a good mentor, it is imperative that you have a good relationship with your mentees. To me, a good mentor-mentee relationship consists of mutual trust, responsibility, amount and reliability of knowledge and information, kindness and sensitivity.
There are many different kinds of OLs, since there are, on average, 700+ OLs for any given August Orientation. Everyone’s values are different, and therefore every student’s orientation experience will be different. Since there is no reputation system or evaluation process for freshmen, “bad” – by which I mean irresponsible and unreliable – OLs will go undetected. Since many OLs re-apply for the position year after year, and are more likely to be chosen based on their prior experience, it is a good idea to have a reputation system.
I think the system should have two parts; feedback from incoming students and feedback from friends. Each member of an OL’s group and several close friends should rate the OL on the categories of trustworthiness, responsibility, knowledgeability, reliability, kindness, and sensitivity. The highest rating should be green, followed in order by blue, yellow, orange, and red. An OL with a rating of green or blue would be a reputable OL; yellow would be neutral; orange and red would be disreputable. Hopefully the institution of a reputation system would force OLs to take their responsibilities more seriously because they would know they are being reported on since, as Resnick et. al. state, "an expectation that people will consider one another’s pasts in future interactions constrains behavior in the present."
In addition, the Orientation Steering Committee members and Orientation Supervisors are selected from previous OLs. As stated by Olson, “if higher degrees of trust can be established, organizations can work more efficiently, and adapt more quickly to changing circumstances” (p. 135). A rating system would make sure that those in charge of running the Orientation Program and those that are in charge of ensuring a positive and nurturing orientation experience for incoming students are truly committed and suited to such a task, and make sure that the best people are consistently selected for the task.
Tuesday, March 31, 2009
Assignment 7 Kevin Singh
An offline community I am involved in is my alumni organization of my high school. All alumni donate money to the organization; thus, a lot of trust is invested in the board members in using this money. Board members could possibly launder the money and use it for vacations. Also, some board members could not be fulfilling their roles in planning alumni events and maintaining a strong relationships to the high school.
Because board members can misuse the donations received we have elections every year. Therefore, a reputation system must be used in order for alumni to judge canadites better. This system should be based on accomplishments of alumni heading/leading projects for the high school such as succesfully petitioning fir more federal funding for the high school. Accomplishments like these allow alumni to make better judgements when choosing board members. The accomplished alumni show leadership, ambition, and care for the organizations goals. These qualities help build bridges of trust between members. The alumni who exhibit these qualities through there accomplishments should be awarded during an award ceremony so that it is made known to everyone in the organization; thereby building a good reputation. When elections occur alumni well be aware of the better reputable alumni. This system could be manipulated because of the awards aspect of the system.
There could be favoritism between members which causes the alumni to select individuals who do not use the donations properly. Elected board members may not fulfill their duties in maintaining good relationships with the high school. Also, they may not plan events properly.
Because board members can misuse the donations received we have elections every year. Therefore, a reputation system must be used in order for alumni to judge canadites better. This system should be based on accomplishments of alumni heading/leading projects for the high school such as succesfully petitioning fir more federal funding for the high school. Accomplishments like these allow alumni to make better judgements when choosing board members. The accomplished alumni show leadership, ambition, and care for the organizations goals. These qualities help build bridges of trust between members. The alumni who exhibit these qualities through there accomplishments should be awarded during an award ceremony so that it is made known to everyone in the organization; thereby building a good reputation. When elections occur alumni well be aware of the better reputable alumni. This system could be manipulated because of the awards aspect of the system.
There could be favoritism between members which causes the alumni to select individuals who do not use the donations properly. Elected board members may not fulfill their duties in maintaining good relationships with the high school. Also, they may not plan events properly.
blog post #7 vincent davis
I am a member of an online community of artists, musicians, and designers called Look Book (http://lookbook.nu). It is an ‘invite only’ website where people can garner free exposure for their clothing design and/or photography and showcase their personal style, essentially a ‘hot or not’ for artists and entertainers where they can post pictures of themselves and the clothes they wear. The Look Book community is actually already maintained through a surprisingly structured reputation system.
Trust is important to LookBook because thousands of people from across the globe put themselves in a vulnerable position when placing a picture of themselves/their outfits in a public forum to be critiqued. The makers of look book wanted the site To “bring together diverse, aesthetically talented and open-minded people around the world and showcase the best in international street fashion” users are encouraged to post their own ‘looks’, but also comment on other users’ style and help vote for which members will appear on the front page of the site at any give time through voting for, or ‘hyping’ that person’s look. Look Book thrives because users are held responsible for their photo-posts, comments, and ratings; the combination of the number of individual posts makes, the number of comments they give/receive on looks, and voting for or ‘hyping’ someone’s post, the combination of all of those factors translates into the users ‘Karma Score’.
A Look Book user is rewarded for their ‘good behavior’ with a good Karma Score, every LOOKBOOK.nu user affects one another's karma equally, and this is how members reward each other for posting looks that are stylish, unique, and interesting. The front page of the site shows the most popular outfits posted throughout the day, the default setting organizes the rankings by combining how many votes or ‘hype’s a look has received with the Karma score of the user who posted it. Being featured on the front page is essentially the goal of every user, it allows one to promote their website/boutique/design firm, etc as well as obtain their 15 minutes of e-‘fame’.
It is very difficult to manipulate the current system because every member’s activity is influenced to a large extent by the actions of other members of the community. A member of the community can only ‘hype’ a particular look once which stops people from artificially ‘hyping’ themselves to the front page.
Trust is important to LookBook because thousands of people from across the globe put themselves in a vulnerable position when placing a picture of themselves/their outfits in a public forum to be critiqued. The makers of look book wanted the site To “bring together diverse, aesthetically talented and open-minded people around the world and showcase the best in international street fashion” users are encouraged to post their own ‘looks’, but also comment on other users’ style and help vote for which members will appear on the front page of the site at any give time through voting for, or ‘hyping’ that person’s look. Look Book thrives because users are held responsible for their photo-posts, comments, and ratings; the combination of the number of individual posts makes, the number of comments they give/receive on looks, and voting for or ‘hyping’ someone’s post, the combination of all of those factors translates into the users ‘Karma Score’.
A Look Book user is rewarded for their ‘good behavior’ with a good Karma Score, every LOOKBOOK.nu user affects one another's karma equally, and this is how members reward each other for posting looks that are stylish, unique, and interesting. The front page of the site shows the most popular outfits posted throughout the day, the default setting organizes the rankings by combining how many votes or ‘hype’s a look has received with the Karma score of the user who posted it. Being featured on the front page is essentially the goal of every user, it allows one to promote their website/boutique/design firm, etc as well as obtain their 15 minutes of e-‘fame’.
It is very difficult to manipulate the current system because every member’s activity is influenced to a large extent by the actions of other members of the community. A member of the community can only ‘hype’ a particular look once which stops people from artificially ‘hyping’ themselves to the front page.
A7 - Online Polling...
While watching some late-night television, I was instructed by TV-host Stephen Colbert to log onto NASA's website and participate in an online voting contest for the name of a new space-node for the International Space Station. The name was, of course, "Colbert" and to my (somewhat) surprise, Stephens followers pushed "Colbert" into the lead. In second place was "Serenity". Almost immediately, NASA considered ignoring this suggestion.
To some the solution is straightforward, it's a stupid name, don't use it. For other, however, it's subverting the will of the people and specifically the will of people who took time out of there day to participate in a vote which they believed would result in the name of a new space station. There is an argument on both sides, so I propose a way that these types of polls can be implemented without such problems. Are real issue is this: When does the poll "trust" the right answer. Here we define trust explicitly as "Not coming from a flood of biased individuals"
As Nasan Bos (Et Al.) point out, research shows that it can be very difficult to develop trust in an online setting, as compared to face-to-face. This is further compounded if the individuals don't think they will interact in the future with other members of the community, or in this case the pollsters. In the case of the NASA poll, both of these issues are troublesome for creating a reputation system. As resnik states reputation systems must have three components
1)Long-lived entities that inspire an expectation of
future interaction;
2)Capture and distribution of feedback about current
interactions (such information must be visible
in the future); and
3)Use of feedback to guide trust decisions.
In this case we can't have at true reputation system, because we can't track users. We can all but give up on 1), unless we envisioned a universal voting website with users etc. and functionality like e-bay.(which is actually a good website concept, but that's another story). The problem we deal with occurs in (2) because the distributions of feedback about the current interaction is skewed, in our case, from individuals biased from TV who then flood the site. This immediately puts all other choices at a disadvantage. To counter-act this sort of flooding-bias, "High rated" names should be put on a "chopping block" and thus each vote actually consists of two voted: "One name to suggest" and "one name to reject." The result allows "down votes", or votes that take away from the total tally of the original voting count. This method could eliminate instances in anonymous voting where someone floods the poll with biased user. This feedback mechanism encapsulates component (3) because we allow feedback from the poll to further enhance the trustworthiness of the population by putting the population in a position to promote any name, and demote any popular name. What do you think?
To some the solution is straightforward, it's a stupid name, don't use it. For other, however, it's subverting the will of the people and specifically the will of people who took time out of there day to participate in a vote which they believed would result in the name of a new space station. There is an argument on both sides, so I propose a way that these types of polls can be implemented without such problems. Are real issue is this: When does the poll "trust" the right answer. Here we define trust explicitly as "Not coming from a flood of biased individuals"
As Nasan Bos (Et Al.) point out, research shows that it can be very difficult to develop trust in an online setting, as compared to face-to-face. This is further compounded if the individuals don't think they will interact in the future with other members of the community, or in this case the pollsters. In the case of the NASA poll, both of these issues are troublesome for creating a reputation system. As resnik states reputation systems must have three components
1)Long-lived entities that inspire an expectation of
future interaction;
2)Capture and distribution of feedback about current
interactions (such information must be visible
in the future); and
3)Use of feedback to guide trust decisions.
In this case we can't have at true reputation system, because we can't track users. We can all but give up on 1), unless we envisioned a universal voting website with users etc. and functionality like e-bay.(which is actually a good website concept, but that's another story). The problem we deal with occurs in (2) because the distributions of feedback about the current interaction is skewed, in our case, from individuals biased from TV who then flood the site. This immediately puts all other choices at a disadvantage. To counter-act this sort of flooding-bias, "High rated" names should be put on a "chopping block" and thus each vote actually consists of two voted: "One name to suggest" and "one name to reject." The result allows "down votes", or votes that take away from the total tally of the original voting count. This method could eliminate instances in anonymous voting where someone floods the poll with biased user. This feedback mechanism encapsulates component (3) because we allow feedback from the poll to further enhance the trustworthiness of the population by putting the population in a position to promote any name, and demote any popular name. What do you think?
Assignment #7 - Austin Lin (akl29)
This semester I am working in a relatively large group on a semester long project for CS 5150 Software Engineering. Defined as defined as a “willingness to be vulnerable, based on positive expectations about the actions of others,“ (Bos, et al. 2002) trust between the 7 team members is very important and is a constant issue that we need to deal with. Because our grades all depend heavily on how much we contribute as well as the overall completion of the project, we need to be able to trust each other with assigned portions of the project, meeting deadlines, attending meetings. There is nobody going back to explicitly check on each person’s contributions so there is a certain level of trust in that everybody contributes high quality work. Completing consistently any of the above would be considered desirable behaviors. Thus completion of the project within deadlines and with quality work reflects cognitive trust. Emotional trust is seen in intergroup relations and the consequences of the course grading.
In an ideal reputation system, each person could receive essentially karma points for contributing over the course of the semester. Though it would be hard to implement and keep a tally of points, a user could gain points by consistently meeting deadlines and attending meetings. Producing good work that is rated by other users would give additional karma points. A reward system could be built in so people who have worked harder on certain areas can be assigned fewer tasks in the future; also people have direct feedback on how much they have given to the project so they know if they need to contribute more. This reputation system would work best in a situation where the ratings would be “portable” in multiple classes throughout a students college years. This would effectively create a “shadow of the future” and affect being chosen for future group projects.
The system could be manipulated in a few ways; first users who realize they have contributed more than average would become worse team members by contributing far less. A decay system could be implemented to keep very high ranked users from abandoning the project thinking that they have contributed enough. Also users who have low rankings due to poor quality work or insufficient ability would be assigned more responsibility, which may lead to a cycle of lowering ratings as well as trust. Collusion between members in which friends help each other can also lead to slightly inflated rankings.
In an ideal reputation system, each person could receive essentially karma points for contributing over the course of the semester. Though it would be hard to implement and keep a tally of points, a user could gain points by consistently meeting deadlines and attending meetings. Producing good work that is rated by other users would give additional karma points. A reward system could be built in so people who have worked harder on certain areas can be assigned fewer tasks in the future; also people have direct feedback on how much they have given to the project so they know if they need to contribute more. This reputation system would work best in a situation where the ratings would be “portable” in multiple classes throughout a students college years. This would effectively create a “shadow of the future” and affect being chosen for future group projects.
The system could be manipulated in a few ways; first users who realize they have contributed more than average would become worse team members by contributing far less. A decay system could be implemented to keep very high ranked users from abandoning the project thinking that they have contributed enough. Also users who have low rankings due to poor quality work or insufficient ability would be assigned more responsibility, which may lead to a cycle of lowering ratings as well as trust. Collusion between members in which friends help each other can also lead to slightly inflated rankings.
Assignment 7: Eric Gunther
I've been following a website for about 5 years now and earlier last year I was finally able to join their online forums. This site has a unique system where the forum users have to be approved by the site webmaster before being let in, so many people have a serious interest vested in this forum. Thus users frequently post and the community has become rather tight-nit, something that doesn't occur a lot on the internet.
A reputation system for this community is already in place, but it contains many aspects that are worth analyzing. When a user makes a comment, to the left of their comment appears their username, avatar, date they joined, number of posts, number of thanks, and a small caption. Underneath the comment appears a custom signature that the user usually puts a banner and some quotes in. This community discusses a lot of things, including technology, entertainment, politics, music and design. Because this is a wide range of topics, only some users may have valid, or trusted opinions in these areas. If I were to post a question about a math problem on the forum, I would want a trusted answer that I know is right. I could verify this by using the reputation system; I would need to see a response from a user with a high number of thanks (everytime someone posts help and you appreciated it, you can give them a "thanks") and perhaps a sign in the signature that shows they are part of a "tech crew" (users create online "crews" that put a sign in the signature to show what crew they're in).
Because this is a forum the most desirable behavior is to have consistent postings and a large, diverse, community. People should be posting answers to questions and their opinions often, thus they can be rewarded by receiving thanks, or invitation to a reputable "crew". This system could be manipulated by a random user because they could always put any crews sign in their signature, but this would probably result in them being defamed by other users. Not to mention the fact that all the users on this forum had to apply to get in, so they already have an interest to maintain a high reputation. In fact, because all the users had to apply to get in and most likely will only have one username it reduces the situation where "good name is not at stake...the temptation to "hit and run" outweighs the incentive to cooperate" (Resnick, 46).
Finally the system rewards users who contribute monetarily towards maintaining server costs (which is an issue for the sharing section of the forum). These users are allowed to have higher resolution avatars and an extra caption box to the left of their posts. While this doesn't seem like a big deal, these users have a very high reputation in the community and their posts are considered much more important than other posts.
A7 Reputation Systems in Slope Media Group
I have dedicated a good amount of time this semester to supporting technology and production with Slope Media Group. This group consists of managers, creative directors, content producers, and techies that work together to report and entertain Cornellians. There are certain key players that have and/or are currently helping Slope Media Group stay alive through the past four years. Establishing a reputation system where positive rep credits could be aggregated might be a clear useful way of establishing a reputation system. This reputation system would be a count of rep credits gained. The rules for gaining rep credits would be established as part of the organizational constitution and agreed upon by all members. Exemplary situations for gaining credits include:
- troubleshooting and fixing a technical radio problem.
- documenting troubleshooting tips for a technical radio problem.
- contributing to the creation of a magazine layout.
- contributing to a bug fix on slopemedia.org
These points would be updated in a central database. Then, each person's profile would indicate their rep credits for the life of the organization. In this way, awareness would be raised about the added value of each member. Of course, it is possible that this system be manipulated just as easy as any hacker would manipulate a vulnerable database. But within the organization, it would occur only with changes to the constitution by a majority member vote.
Assignment #7:Trust in a Club (Abena Oteng-Agipong)
For the last 3 years, I have been part of the Cornell Wushu club; this year I am on the executive board as Treasurer. As a member of the executive board, I require the trust of the rest of the board and team. In return, they need to be able to trust in me (that I won't misuse the money or steal it). In addition, because we perform a lot of demonstrations throughout the year, we need to trust in each other in order to organize a performance and teach each other different techniques. Just as trust is important for businesses to collaborate, trust is needed for clubs to collaborate and essentially function.
With a high degree of trust established, “organizations can work more efficiently, and adapt more quickly to changing circumstances”(Bos 135). This is very important for my club because the circumstances and space in which we perform changes drastically from one performance to the next. The desire behavior for our club is rather simple; pretty much, a person should attend club practices regularly. That way, the rest of the team, who have a “known past histories or the prospect of future interaction”(Resnick) can get to the know to person and start building trust with that person. In turn, that person will end up learning more basic forms, techniques and eventually weapons forms. The biggest reward system would definitely be getting to learn one or more weapon forms.
Our group doesn’t suffer too much from intervals of delayed or fragile trust because my club is a FtF community; senior members (as well as other members) can pick up non-verbal cues required to make judgments about one’s trustfulness. However, a way our scheme could be manipulated is if a senior member had a secret agenda and for some reason, taught other members of the group the wrong technique. Also, if someone in charge of leading some of the practices all of a sudden stops attending all practices, the rate in which the team learns techniques and moves to be used in demos decreases dramatically. Lastly, misuse of resources by the executive board greatly reduces trust within the team.
With a high degree of trust established, “organizations can work more efficiently, and adapt more quickly to changing circumstances”(Bos 135). This is very important for my club because the circumstances and space in which we perform changes drastically from one performance to the next. The desire behavior for our club is rather simple; pretty much, a person should attend club practices regularly. That way, the rest of the team, who have a “known past histories or the prospect of future interaction”(Resnick) can get to the know to person and start building trust with that person. In turn, that person will end up learning more basic forms, techniques and eventually weapons forms. The biggest reward system would definitely be getting to learn one or more weapon forms.
Our group doesn’t suffer too much from intervals of delayed or fragile trust because my club is a FtF community; senior members (as well as other members) can pick up non-verbal cues required to make judgments about one’s trustfulness. However, a way our scheme could be manipulated is if a senior member had a secret agenda and for some reason, taught other members of the group the wrong technique. Also, if someone in charge of leading some of the practices all of a sudden stops attending all practices, the rate in which the team learns techniques and moves to be used in demos decreases dramatically. Lastly, misuse of resources by the executive board greatly reduces trust within the team.
Assignment #7- Jordan Meltzer
I am a member of an online forum for the television show, Lost. There are many members, but only a handful of Moderators. Members trust that moderators will review posts and remove forms of trolling or baiting, when users engage in personally directed insults or attacks to elicit inflammatory posts. Moderators may suspend or ban members who engage in such activity, and other members trust that rules are enforced. Trust may be defined as “willingness to be vulnerable, based on positive expectations about the actions of others” (Bos et al, 135). For example, if I post a question indicating that I do not understand something in an episode, I am putting myself in a particularly vulnerable state. I am trusting others to give thoughtful responses and not complain that I should pay more close attention.
In addition to treating others with respect in responding to posts, another desired behavior is to place posts in appropriate threads. For example, if a member is a few episodes behind and wishes to discuss the events of a certain episode, they should not be given any information from newer episodes. It is encouraged that episode-specific posts are placed in their particular thread and any revealing information is placed in spoiler tags (spoiler tags must be highlighted to see message). Thus, it is a desired behavior for members to understand and respect purposes of each thread. Meeting desired behaviors is rewarded by receiving thoughtful responses and increasing post count. Resnick et al (2000) state that reputation systems distribute feedback about past behavior and are designed to promote trust and keep users engaged. The main source of feedback is one’s post count and status, since all posts are monitored and if posts do not violate rules or expectations then they are left as is, increasing post count. By obtaining higher post count, members can increase status (junior /standard/advanced/senior member). There is delayed trust associated with status, such that moderators are more likely to monitor posts of lower-ranked members, since “CMC may delay trust formation by slowing the rate at which individuals can gather nonverbal cues about partners’ trustworthiness” (Bos et al, 137). The online nature of the forum may make it seem members are detached from Moderators, so “good name is not at stake…the temptation to ‘hit and run’ outweighs the incentive to cooperate” (Resnick et al, 46). However, the first five posts of every member are not publicly displayed until they have been read through and approved by Moderators. This helps keep users from trolling, since they must make five acceptable, Lost-relevant posts before they can even directly post to the forum. If a user really wishes to manipulate the system, then they could make five relevant posts and then create as many spamming, inflammatory messages as possible before they are caught, in which case they would be permanently banned from the online forum.
In addition to treating others with respect in responding to posts, another desired behavior is to place posts in appropriate threads. For example, if a member is a few episodes behind and wishes to discuss the events of a certain episode, they should not be given any information from newer episodes. It is encouraged that episode-specific posts are placed in their particular thread and any revealing information is placed in spoiler tags (spoiler tags must be highlighted to see message). Thus, it is a desired behavior for members to understand and respect purposes of each thread. Meeting desired behaviors is rewarded by receiving thoughtful responses and increasing post count. Resnick et al (2000) state that reputation systems distribute feedback about past behavior and are designed to promote trust and keep users engaged. The main source of feedback is one’s post count and status, since all posts are monitored and if posts do not violate rules or expectations then they are left as is, increasing post count. By obtaining higher post count, members can increase status (junior /standard/advanced/senior member). There is delayed trust associated with status, such that moderators are more likely to monitor posts of lower-ranked members, since “CMC may delay trust formation by slowing the rate at which individuals can gather nonverbal cues about partners’ trustworthiness” (Bos et al, 137). The online nature of the forum may make it seem members are detached from Moderators, so “good name is not at stake…the temptation to ‘hit and run’ outweighs the incentive to cooperate” (Resnick et al, 46). However, the first five posts of every member are not publicly displayed until they have been read through and approved by Moderators. This helps keep users from trolling, since they must make five acceptable, Lost-relevant posts before they can even directly post to the forum. If a user really wishes to manipulate the system, then they could make five relevant posts and then create as many spamming, inflammatory messages as possible before they are caught, in which case they would be permanently banned from the online forum.
Assignment 7: BONESAW IS READY! (Kyle Barron)
A community in which I consider myself an active member is the Bowling Club. Every Monday is league play, which is competitive in that we keep score, but everyone still has fun. Helen Newman lanes is still the only bowling alley I have seen to this day that does not use electronic scoring. All the scoring is done by hand.... human hand... The vice president of the club is in charge of collecting the scorecards, compiling the scores for the day, then posting the week's results via the bowling club e-list. If I hadn't made it clear before, I would like to restate that the lanes are quite old. Sometimes the sweep feels like clearing the pins after one roll. Although it may sound peculiar, it happens more often than one would think. If this is the situation, instead of bothering the lane staff who have better things to do than set up pins by hand at a lane with automatic pin-setting, we just let whoever is bowling aim at the spot to pick up the spare. If we judge that it was good, then we'll give them the spare. I've outlined three scenarios in which trust matters. If we did not trust, then the club would be having much less fun.
Let's look at the first situation: scoring by hand. Each week, any given team of three will play another team of three for three games. At the end of the games, the captains from each time sign the scorecard to verify that the other team's scores are correct. I have not seen any scorecards signed because people trust each other. The vice president could alter the scores, but she chooses not to because it only takes one person to notice their score has been recorded improperly. Right now, there is a pretty solid system of checks and balances.
A system we could implement would be to install electronic scoring, but that's not exciting. A system still involving people would be to have the captains actually sign the scorecards. This will at least make them take a glance at the sheet to make sure no one put down 300 for their score for one game. This scheme is unfortunately susceptible to corruption, bribery, and interactions of the like. A better scheme would be to establish an average at lanes with electronic scoring, so that when they return to Helen Newman, the whole league will know what is considered average.
Assignment 7: John Fox
The online community that I have been using the most lately is my XBOX Live community. As an owner of a XBOX 360 for the past year, I have had access to gamers all over the world who have the common interest in gaming. The amount of trust needed in this online community usually depends on the type of games that you may be playing. There is a certain level of trust needed for each game but the reason changes from game to game. In a game like Call of Duty you are going to want someone on your team, who is going to have your back, who is focused on the goal at hand. Playing against someone in a sports game you want someone who wont cheat or quit if they are playing badly. That's why they have a system to add friends. If they are your friends in real life you can hold them accountable, and if you have an enjoyable experience playing with someone even though you don't know them you can add them because the trust has had time to develop through experience.
Desirable behavior would be any behavior that allows for smooth gameplay when playing online. Players that don't curse consistently during play, that don't quit in the middle of games, or repeatedly sabotage teams during team play are people who get rewarded. Resnick's idea of "reputation systems" is utilized in the creation of the XBOX Live system. By having players file reports after games for misconduct or giving people high ratings after an enjoyable game, you create a system where past interactions can have weight and influence on future interactions in the community. Obviously when it comes to rating systems, there are a lot of gray areas when it comes down to the truthfulness of the information. If you had a group of friends that play games against each other and give each other good ratings regardless of what happened in the game, to the public it may seem like they are gamers that you would want to interact with. In reality they may be the very opposite. When control of information is outside of the administrations hands it is easy to manipulate the results. XBOX Live tries to provide more information by having a rating system that measures success based on tasks done. It may not tell a complete story but it helps users define potentially good competition and users who believe in fair play.
Desirable behavior would be any behavior that allows for smooth gameplay when playing online. Players that don't curse consistently during play, that don't quit in the middle of games, or repeatedly sabotage teams during team play are people who get rewarded. Resnick's idea of "reputation systems" is utilized in the creation of the XBOX Live system. By having players file reports after games for misconduct or giving people high ratings after an enjoyable game, you create a system where past interactions can have weight and influence on future interactions in the community. Obviously when it comes to rating systems, there are a lot of gray areas when it comes down to the truthfulness of the information. If you had a group of friends that play games against each other and give each other good ratings regardless of what happened in the game, to the public it may seem like they are gamers that you would want to interact with. In reality they may be the very opposite. When control of information is outside of the administrations hands it is easy to manipulate the results. XBOX Live tries to provide more information by having a rating system that measures success based on tasks done. It may not tell a complete story but it helps users define potentially good competition and users who believe in fair play.
Assignment #7 - Peter Clain
Fraternities tend to be fairly structured organizations, but aside from any outward presentations or special sets of rules, a fraternity is simply a large collection of people living together. My fraternity is no different, and like any group that relies on each of its members, trust plays an important role in our ability to accomplish all the tasks that help keep our organization together.
In dealing with these tasks, work is distributed among each of the members. Members need to complete house jobs, attend meetings, and maintain responsibility for any leadership positions they may have in the house. Trust between members is so important because of this reliance on one another. The person in charge of ordering food, for example, requires a huge amount of trust from each of the members. Without proper attention, there would be nothing to eat or drink, but by consolidating this task to one person, the organization can run more efficiently. With higher degrees of trust, “organizations can work more efficiently and adapt more quickly to changing circumstances” (Bos, Olson…).
Desirable behavior from a member includes the completion of each of their tasks, but because a fraternity can have so many tasks/members, it is often difficult to blame someone when a task is not completed. A reputation system would help by rewarding dependable members and making desirable/undesirable behaviors more public. This way, tasks with more responsibility could be assigned to more dependable members, and less dependable members would be encouraged to improve their behavior. The expectation that past actions would affect future interactions would encourage everyone to succeed.
The reputation system would work by rewarding/punishing members with a visible form of feedback. Members with a higher reputation could get access to certain assets, perhaps, or they could receive special clothing to act as a status symbol. Less dependable members would receive the opposite treatment. Reputation would be measured by task completion and attendance at weekly meetings, and members at the meetings would reach a consensus regarding each member’s reputation (which would also increase attendance).
While this scheme helps provide feedback, it could be manipulated by avoiding blame through inactivity or by collaborating with friends in the fraternity. Members are usually friendly with one another, by nature. In addition, no matter what the rewards would be, having a high reputation would most likely be seen as “elite,” and members would manipulate the system to achieve it.
In dealing with these tasks, work is distributed among each of the members. Members need to complete house jobs, attend meetings, and maintain responsibility for any leadership positions they may have in the house. Trust between members is so important because of this reliance on one another. The person in charge of ordering food, for example, requires a huge amount of trust from each of the members. Without proper attention, there would be nothing to eat or drink, but by consolidating this task to one person, the organization can run more efficiently. With higher degrees of trust, “organizations can work more efficiently and adapt more quickly to changing circumstances” (Bos, Olson…).
Desirable behavior from a member includes the completion of each of their tasks, but because a fraternity can have so many tasks/members, it is often difficult to blame someone when a task is not completed. A reputation system would help by rewarding dependable members and making desirable/undesirable behaviors more public. This way, tasks with more responsibility could be assigned to more dependable members, and less dependable members would be encouraged to improve their behavior. The expectation that past actions would affect future interactions would encourage everyone to succeed.
The reputation system would work by rewarding/punishing members with a visible form of feedback. Members with a higher reputation could get access to certain assets, perhaps, or they could receive special clothing to act as a status symbol. Less dependable members would receive the opposite treatment. Reputation would be measured by task completion and attendance at weekly meetings, and members at the meetings would reach a consensus regarding each member’s reputation (which would also increase attendance).
While this scheme helps provide feedback, it could be manipulated by avoiding blame through inactivity or by collaborating with friends in the fraternity. Members are usually friendly with one another, by nature. In addition, no matter what the rewards would be, having a high reputation would most likely be seen as “elite,” and members would manipulate the system to achieve it.
Assignment 7: Tom Ternquist
An online community that has an interesting use of trust in its system is Twitter. What makes it so interesting is how simple the system in terms of its basic functionality. That being said, its use has exploded, and in ways that are far from simple. Twitter has become the ultimate technology for the dissemination of information. Whether it’s for getting the word out about a political campaign, spreading breaking news, or if it’s used in a viral marketing campaign, Twitter is unmatched in its ability to get information to its users.
Trust becomes of great importance at several levels, the most obvious being the notion that a user is trusting of the other users who follow them on the service are not going to use the user’s information for harm. But to me, what is the most interesting is the system of trust that forms around the idea of who is an authority for a given subject on Twitter.
Users can essentially perform their own data mining in their searches for users to follow on any given subject. However, it’s clear that some users are more likely to provide a higher level of quality information than others. A user may claim to be a political science aficionado, but how can we really be confident of this?
A rather elegant and simple solution is to follow an approach similar to Kleinberg’s Hubs and Authorities model, but simply use a user’s followers and friends as a metric for importance. If Twitter or an external service through the Twitter API implements this system, it would serve as self-reinforcing mechanism that should identify who are the authorities on Twitter. Users who produce good content should gain more followers, and from this, the system rewards the user by boosting their authority. A higher authority should give the user access to more followers, and this process should continue until some equilibrium is reached. To an extent, this process happens naturally through name recognition and other marketing, but having an automated system would certainly help to identify these authorities.
Of course the scheme can be manipulated. An obvious way for this to happen is that users who are authorities in one subject area can use their popularity to claim authority in some other domain. However, I would suspect that users who are highly respected would not try to risk their reputation by manipulating the system.
Trust becomes of great importance at several levels, the most obvious being the notion that a user is trusting of the other users who follow them on the service are not going to use the user’s information for harm. But to me, what is the most interesting is the system of trust that forms around the idea of who is an authority for a given subject on Twitter.
Users can essentially perform their own data mining in their searches for users to follow on any given subject. However, it’s clear that some users are more likely to provide a higher level of quality information than others. A user may claim to be a political science aficionado, but how can we really be confident of this?
A rather elegant and simple solution is to follow an approach similar to Kleinberg’s Hubs and Authorities model, but simply use a user’s followers and friends as a metric for importance. If Twitter or an external service through the Twitter API implements this system, it would serve as self-reinforcing mechanism that should identify who are the authorities on Twitter. Users who produce good content should gain more followers, and from this, the system rewards the user by boosting their authority. A higher authority should give the user access to more followers, and this process should continue until some equilibrium is reached. To an extent, this process happens naturally through name recognition and other marketing, but having an automated system would certainly help to identify these authorities.
Of course the scheme can be manipulated. An obvious way for this to happen is that users who are authorities in one subject area can use their popularity to claim authority in some other domain. However, I would suspect that users who are highly respected would not try to risk their reputation by manipulating the system.
Assignment 7: Ashley Vernon
An online community that I am a part of is FanFiction.net. FanFiction is fictional stories about a show that you are a fan of. These are also known as fanfics. FanFiction.net is an archive and forum where fanfic readers and writers around the globe share their passion.
According to Resnick, "a reputation system collects, distributes, and aggregates feedback about participants' past behavior". Based on this definition, I don't think FanFiction.net has that great of a reputation system. Currently, the only information it displays about a member's past behavior is the number of fanfics they have written, and how long they have been a member of the site. It also displays their reviews of fanfics, but these are spread out all through the site, and difficult to find. I think their reputation system should be updated because it doesn’t help people decide who to trust, or encourages trustworthy behavior.
Resnick has defined 3 properties of reputation systems: (1) long-lived entities that inspire an expectation of future interaction (2) capture and distribution of feedback about current interactions that will be visible in the future (3) use of feedback to guide trust decisions. According to this, my reputation system for the community would consist of members receiving levels for their behavior, such as an enthusiast. I think that his would help people decide who to trust, as trust plays a role in this community. Members trust that the authors actually wrote the stories themselves, and are not plagiarizing someone else's work. Members also trust that other members are not going to “flame” their stories in a review. The desirable behaviors in this community are to establish continuing relationships with other members, write creative fanfics, and leave well rounded critiques in reviews of fanfics.
My reputation system would include a member’s profile including not only how long they’ve been a member and how many stories they’ve written, but also if they are a frequent contributor, and if they are making significant contributions. This scheme could be manipulated because users could just review everyone’s story and not really say anything significant. To prevent this, the author of the story should able to indicate how helpful each review of their story was. Other members should also be able to indicate if the review was good, possibly by clicking on a “thumbs up” or a “thumbs down”. I think this would deter flaming. I would reward members by highlighting top-quality contributors. For example, if one member always leaves very helpful reviews that are well thought out, they would be acknowledged.
According to Resnick, "a reputation system collects, distributes, and aggregates feedback about participants' past behavior". Based on this definition, I don't think FanFiction.net has that great of a reputation system. Currently, the only information it displays about a member's past behavior is the number of fanfics they have written, and how long they have been a member of the site. It also displays their reviews of fanfics, but these are spread out all through the site, and difficult to find. I think their reputation system should be updated because it doesn’t help people decide who to trust, or encourages trustworthy behavior.
Resnick has defined 3 properties of reputation systems: (1) long-lived entities that inspire an expectation of future interaction (2) capture and distribution of feedback about current interactions that will be visible in the future (3) use of feedback to guide trust decisions. According to this, my reputation system for the community would consist of members receiving levels for their behavior, such as an enthusiast. I think that his would help people decide who to trust, as trust plays a role in this community. Members trust that the authors actually wrote the stories themselves, and are not plagiarizing someone else's work. Members also trust that other members are not going to “flame” their stories in a review. The desirable behaviors in this community are to establish continuing relationships with other members, write creative fanfics, and leave well rounded critiques in reviews of fanfics.
My reputation system would include a member’s profile including not only how long they’ve been a member and how many stories they’ve written, but also if they are a frequent contributor, and if they are making significant contributions. This scheme could be manipulated because users could just review everyone’s story and not really say anything significant. To prevent this, the author of the story should able to indicate how helpful each review of their story was. Other members should also be able to indicate if the review was good, possibly by clicking on a “thumbs up” or a “thumbs down”. I think this would deter flaming. I would reward members by highlighting top-quality contributors. For example, if one member always leaves very helpful reviews that are well thought out, they would be acknowledged.
Assignment #7: reputation systems (Lisa Park)
A community that I am a part of is my group of three apartment-mates. Trust is very important to this community because we share living spaces that we trust each other to take care of. In the article by Bos, et. all, "Trust is best defined as a 'willingness to be vulnerable, based on positive expectations about the actions of others'". As stated in the Resnick article, trust is built up over time, based on experiences you have had with an individual. It is this information about the past that enables one to (or deters one from) trusting another. It is also the awareness that your actions now will affect how someone interacts (or doesn't) with you in the future that makes you do trustworthy things. Thus, in a shared living situation, we build trust in one another by being responsible about shared resources: keep kitchen clean, don't hog the bathroom, take out the trash and respecting personal spaces: turn the TV down, don't eat someone else's food, pay people back. Without this trust, we may "avoid collaborating with others altogether" (Bos, 135), providing for a tense and unenjoyable living situation.
Because we can't be constantly watching over each other, we need a way to make our merits/non-merits visible to allow for the trust-building process. This is where a reputation system would come in handy. One possible accountability system could be based on reports posted on the fridge. Every time a roomie did something for the group (take out trash, scoop cat's litter, buy more milk), they could write it on the fridge, under their name. Every time someone had a problem with another roomie's behavior, they could write this under the offending roomie's name. The incentive for good behavior is reputation, as you'd want to avoid having an empty "good" list and you'd want recognition for the contributions you've made. The "grievances" lists are needed because someone may still have a complaint about your performance that you weren't even aware of.
Possibly, one could manipulate the system by taking credit for something they didn't do, though this would probably be obvious. Others may post unfair or uncalled for complaints on your sheet, marring your reputation. One could also hurt themselves by not posting about their good deeds.
Because we can't be constantly watching over each other, we need a way to make our merits/non-merits visible to allow for the trust-building process. This is where a reputation system would come in handy. One possible accountability system could be based on reports posted on the fridge. Every time a roomie did something for the group (take out trash, scoop cat's litter, buy more milk), they could write it on the fridge, under their name. Every time someone had a problem with another roomie's behavior, they could write this under the offending roomie's name. The incentive for good behavior is reputation, as you'd want to avoid having an empty "good" list and you'd want recognition for the contributions you've made. The "grievances" lists are needed because someone may still have a complaint about your performance that you weren't even aware of.
Possibly, one could manipulate the system by taking credit for something they didn't do, though this would probably be obvious. Others may post unfair or uncalled for complaints on your sheet, marring your reputation. One could also hurt themselves by not posting about their good deeds.
Assignment 7 - Reputation - Jesse Miner
Our INFO 3650 class - specifically the blog - can be considered a community. Trust is important in this context because it is good to know a little bit about the author and his or her previous posts when you sit down to read the blog. ("Is this post likely to be good, or am I wasting my time by reading it?") Desirable behaviors for blog posts include writing interesting entries that fulfill the requirements of the assignment. Good posts are also well-written and do not include unnecessary or inappropriate information. The main reward (and motivation) for good behavior is a high grade, but a reputation system could have other rewards, too. For example, there could be a rating system in which students could rate others' posts, maybe on a scale from 1 to 5. Since people might not want ratings displayed next to their posts, there could be an option to view the highest-rated posts without seeing the actual scores. Similarly, one could search for users whose posts were rated highly to find reliable authors. Each user could have an overall reputation score equal to the average of all of his or her posts' ratings. This overall score would be visible to other people.
This rating system could help the instructors pick out the best blog posts each week and share them in class as an additional reward (with the author's consent, of course). However, the ratings probably shouldn't be visible to the instructors until after the grading is done to avoid bias.
A weakness of this system is that it could easily be manipulated, especially since many members of the class know each other. People could agree to give each other's posts high ratings to inflate their reputations. Then honest people who did not collude would be at a disadvantage, even if they worked hard and contributed high-quality posts. Also, the system might have the negative side effect of discouraging people who receive low ratings on their posts. On the other hand, they might be inspired to try harder to earn a better reputation.
This rating system could help the instructors pick out the best blog posts each week and share them in class as an additional reward (with the author's consent, of course). However, the ratings probably shouldn't be visible to the instructors until after the grading is done to avoid bias.
A weakness of this system is that it could easily be manipulated, especially since many members of the class know each other. People could agree to give each other's posts high ratings to inflate their reputations. Then honest people who did not collude would be at a disadvantage, even if they worked hard and contributed high-quality posts. Also, the system might have the negative side effect of discouraging people who receive low ratings on their posts. On the other hand, they might be inspired to try harder to earn a better reputation.
Assignment 7: Michael Triche
National Society of Black Engineers(NSBE) is the community offline that I will use for this assignment. The organization is student run and comprises of over 25,000 members. Trust matters greatly in this community especially being a large student run organization where each member depends on the other to do their part to maintain success. These include paying dues, helping spread the message of NSBE, attending conferences, applying to internships and jobs, keeping grades up, and doing their share of the work to make each chapter run smoothly amongst others. There has to be trust that the chairman and the elected officials of the various boards will do there jobs and work with the best interest of NSBE at hand. Each person is vulnerable to a decision of everyone else, if the chairman quits then that would effect the whole organization. The relationships built become more than business relationships, they become "emotional trust" since everyone has to work together collectively and be the support system of one another in order to get work done. There is also cognitive trust in the capabilities of others counterparts.
The desired behavior is being a hard worker and trust worthy. Most of the chairmans and other officials elected have to have a good reputation from his or her peers much like the rating reputation system for Ebay. Resnick shows this when he states "isolated interactions take attributes of long term relationships". Alot of NSBE members began while in high school so over the years they connect and begin to see those who distinguish themselves as leaders in the organization who are trustworthy. The members who are the most trust worthy or respected win elections and receive awards based on their dedication and hardwork. some loop holes to the system is if people only vote on popularity and not the acheivements of those running for positions. As Resnick elaborates on, the more information known about someone, the more they could be trusted. Those individuals with the best reputations get the most accolades. These reputations have to be as Resnick states based on ensured honest reports. Members would not like being in the organization if the chairman only wanted to connect people from his college with employers and thats it.
The desired behavior is being a hard worker and trust worthy. Most of the chairmans and other officials elected have to have a good reputation from his or her peers much like the rating reputation system for Ebay. Resnick shows this when he states "isolated interactions take attributes of long term relationships". Alot of NSBE members began while in high school so over the years they connect and begin to see those who distinguish themselves as leaders in the organization who are trustworthy. The members who are the most trust worthy or respected win elections and receive awards based on their dedication and hardwork. some loop holes to the system is if people only vote on popularity and not the acheivements of those running for positions. As Resnick elaborates on, the more information known about someone, the more they could be trusted. Those individuals with the best reputations get the most accolades. These reputations have to be as Resnick states based on ensured honest reports. Members would not like being in the organization if the chairman only wanted to connect people from his college with employers and thats it.
Assignment 7: Reputation System, Lisa Ji
I realize that I’m always blogging about taiko (Japanese drumming), but it’s my only organized non-academic activity and the group dynamics make it very relevant to the topic of reputation building. Our group is very similar to a sports team, where we must give large amounts of time to practice and performance, but there is less formal motivation to dedicate to the group’s effort since it is just a club, not a university-organized team. As such, the members of taiko must find their own motivation to attend meetings, go to practice, and learn new pieces on their own. However, team members cannot always find this self-motivation, and this is an area where a reputation system could be implemented.
Resnick says that “when people interact with one another over time, the history of past interactions informs them about their abilities and dispositions.” This is certainly true of any group, including taiko, but this information is hard to quantify or objectify. As in, the team does not have a base measure with which it can compare its members. A reputation system could be this baseline. In taiko, group members could earn small amounts of reputation points for daily and weekly activities such as attending practice, getting there early to set up drums, or staying a few minutes at the end to put drums away. Larger numbers of points can be awarded for less frequent events such as general body meetings. Skill can be rewarded by awarding a sum of points for each song in which a person participates in a performance, and enthusiasm can be rewarded by awarding points for attending performances even when not performing. Finally, these points would decay over time, so new members are not forever trying to catch up to older members in point totals.
With this system of point awards in place, rewards would be qualification for performance positions or executive board nominations. This would ensure that people in these positions would have both the skill and dedication to perform well. However, in any reputation system, there is the opportunity to manipulate the scheme. If a trusted member lies about his efforts at practices, or says that he attended a performance but sat away from the group when he did not actually attend, the system falls apart. Members could artificially inflate their reputations to achieve the performance or e-board qualification, yet not deserve the trust that is placed in a person appointed to these positions. Finally, there is the simple matter of getting the system started. Resnick notes that “in the offline world, capturing and distributing feedback is costly.” The extra effort of filling in your points for a day or checking other members’ points is minimal, but still difficult to overcome. These problems could be found in any reputation system, yet the scheme (if successful) could benefit taiko just as much as any community.
Resnick says that “when people interact with one another over time, the history of past interactions informs them about their abilities and dispositions.” This is certainly true of any group, including taiko, but this information is hard to quantify or objectify. As in, the team does not have a base measure with which it can compare its members. A reputation system could be this baseline. In taiko, group members could earn small amounts of reputation points for daily and weekly activities such as attending practice, getting there early to set up drums, or staying a few minutes at the end to put drums away. Larger numbers of points can be awarded for less frequent events such as general body meetings. Skill can be rewarded by awarding a sum of points for each song in which a person participates in a performance, and enthusiasm can be rewarded by awarding points for attending performances even when not performing. Finally, these points would decay over time, so new members are not forever trying to catch up to older members in point totals.
With this system of point awards in place, rewards would be qualification for performance positions or executive board nominations. This would ensure that people in these positions would have both the skill and dedication to perform well. However, in any reputation system, there is the opportunity to manipulate the scheme. If a trusted member lies about his efforts at practices, or says that he attended a performance but sat away from the group when he did not actually attend, the system falls apart. Members could artificially inflate their reputations to achieve the performance or e-board qualification, yet not deserve the trust that is placed in a person appointed to these positions. Finally, there is the simple matter of getting the system started. Resnick notes that “in the offline world, capturing and distributing feedback is costly.” The extra effort of filling in your points for a day or checking other members’ points is minimal, but still difficult to overcome. These problems could be found in any reputation system, yet the scheme (if successful) could benefit taiko just as much as any community.
Assignment 7 - Peter Hunt
I am part of a music group that is rapidly expanding its team. What began as a small group of friends playing in a basement has expanded to a nationally touring band complete with a manager, tour manager, publicist, and lawyer. In this community, trust is vitally important. How are we to know if work is getting done, or money is being properly handled? Who is reliable for what? Trust is especially important because most members of the group are remotely located. As discussed in class, it is much more difficult to trust people who are remotely located.
Desirable behaviors in this community mostly revolve around fulfilling one’s stated responsibilities. Interestingly enough, equal division of labor does not matter much in this community, but fulfilling tasks one accepts in a timely and complete matter does. For example, one builds a reputation by volunteering to complete tasks and completing them well, such as sending press kits out to labels, or driving to a show, etc etc. Eventually, these people prove that they are the “real deal” and are able to be entrusted with real responsibility.
The rewards in this community are few and far between. The music industry has very little money left in it, and until very recently we did not even break even. Monetary rewards are out of the question, but there is a level of respect given to those who do the brunt of the work. When it comes to better sleeping accommodations, or meeting with top industry folks, those who put more in are higher in the pecking order and are able to participate.
This system is so basic, it is difficult to manipulate. Since it is based on tangible, real-world results, one would have to be quite lucky to end up with a higher level of trust than they deserve. Having a results-oriented reputation system like we have ensures the success of the venture no matter the method.
Desirable behaviors in this community mostly revolve around fulfilling one’s stated responsibilities. Interestingly enough, equal division of labor does not matter much in this community, but fulfilling tasks one accepts in a timely and complete matter does. For example, one builds a reputation by volunteering to complete tasks and completing them well, such as sending press kits out to labels, or driving to a show, etc etc. Eventually, these people prove that they are the “real deal” and are able to be entrusted with real responsibility.
The rewards in this community are few and far between. The music industry has very little money left in it, and until very recently we did not even break even. Monetary rewards are out of the question, but there is a level of respect given to those who do the brunt of the work. When it comes to better sleeping accommodations, or meeting with top industry folks, those who put more in are higher in the pecking order and are able to participate.
This system is so basic, it is difficult to manipulate. Since it is based on tangible, real-world results, one would have to be quite lucky to end up with a higher level of trust than they deserve. Having a results-oriented reputation system like we have ensures the success of the venture no matter the method.
Assignment 7 Julie Bai
I am working with a group for the final project in Museums and Public Spheres. Even though there is a deadline for the project, the professor gives a lot of flexibility in regards to the time and depth of content the group can handle. There are no evaluations or submission of progress report to give the group incentives to contribute to the project regularly.
A reputation system would be useful to motivate the group to contribute to the project. According to Resnick, reputation system needs people expecting future interaction and feedback that is collected, distributed and aggregated to guide trust decisions. Trust is important in the group because it helps the group to work together effectively. Without trust, members would be uncertain about other members’ performance, and they would be unwilling to cooperate and communicate effectively to produce quality results.
The reputation system will build and encourage trust among the group by “establishing the shadow of the future (Resnick, 2000).” Shadow of the future explains that expectation of reciprocity of retaliation in future interactions creates an incentive for good behavior (Resnick, 2000). The reputation system will be a forum that allows members to rate each other’s posts (number of stars) and collects points for each post. One’s stars and points will be counted towards the grade for the final project. Ultimately, the system will help to know each other’s abilities, dispositions, and current contribution to the project, and encourage members to be more active in the final project.
If the professor does not look into the forum for content and quality, but just depend on the stars and points members are accruing for the final project to grade, members can collaborate and rate one another positively, artificially inflating their individual contribution.
A reputation system would be useful to motivate the group to contribute to the project. According to Resnick, reputation system needs people expecting future interaction and feedback that is collected, distributed and aggregated to guide trust decisions. Trust is important in the group because it helps the group to work together effectively. Without trust, members would be uncertain about other members’ performance, and they would be unwilling to cooperate and communicate effectively to produce quality results.
The reputation system will build and encourage trust among the group by “establishing the shadow of the future (Resnick, 2000).” Shadow of the future explains that expectation of reciprocity of retaliation in future interactions creates an incentive for good behavior (Resnick, 2000). The reputation system will be a forum that allows members to rate each other’s posts (number of stars) and collects points for each post. One’s stars and points will be counted towards the grade for the final project. Ultimately, the system will help to know each other’s abilities, dispositions, and current contribution to the project, and encourage members to be more active in the final project.
If the professor does not look into the forum for content and quality, but just depend on the stars and points members are accruing for the final project to grade, members can collaborate and rate one another positively, artificially inflating their individual contribution.
Assignment 7: Jeanette D. Pineiro
I recently joined the planning committee for a Global Development conference that is taking place in Cornell next fall. This offline community shares several things in common, such as being interested in global development issues and doing what we can to help. The reputation system for the planning committee can be similar to that of a class. When you attend meetings it is noted and helps build a member’s reputation. Trust is essential in this community because each member is being trusted to complete the jobs or tasks that are assigned to them. If one member fails to do this then it may slow down the group as a whole. Often tasks are interconnected so if one group is falling behind then it will cause a domino effect for other groups, which could compromise the conference in the long run. Desirable behaviors include getting your job done in a timely manner, helping others if they need help, asking for help if you need it yourself, and attending a good amount of meetings. Members are rewarded because they are trusted with more responsibilities and can play a bigger part in the preparation of the conference. However, people can manipulate this by stealing other members’ ideas and claiming it as their own. Also they may miss meetings and claim it is because they have work or are sick when in fact you really have no way of knowing for sure if that is true or not.
Assignment 7 Reputation-Steve Swigut
In my business fraternity trust and reputation are very important to our members. Trust is important because we need to know that we can count on each other for advice, guidance, and help in numerous situations. We often share personal details about our lives and other business information that could be used against us if shared with others outside the fraternity. Reputation helps build trust and therefore we have a reputation system in place setting values on certain events.
We give certain points out for attending/creating professional events, social events, community service events, as well as for helping brothers in need during certain difficult times. Basically, our desirable behaviors is for every member of the fraternity to be as actively involved with the fraternity as possible through both building cognitive and emotional trust. And for our fraternity to become closer and stronger as a result of this build in trust and reputation.
Points are given out at every chapter with various amounts allowed for certain standardized events. Brothers can also suggest giving points to other brothers for certain good deeds either to another brother or the fraternity as a whole. The individual explains why they deserve to be rewarded and the brother being rewarded explains in further detail what he/she did. Then the entire fraternity votes on whether this is to be awarded or not.
Brothers are then rewarded at different times throughout the semester. Brother recognition occurs weekly at chapter and at our banquet each semester. These have only value intrinsically yet reputation and praise from the whole brotherhood is valued by all of us. For more concrete rewards “familys” within our fraternity with the highest total points each semester get the first picks for “littles” who are new fraternity members who will join that family. Family pride is also a driving force in working towards this goal. As Resnick, et. al. explain in “Reputation Systems,” a good design will cause users to expect past interactions to influence future interactions. Past interactions with others within our fraternity builds a reputation and trust within the group allowing for others to understand and predict their future behavior.
Our scheme could be manipulated if our ties as a whole were not strong. One could collude to give each other points as to make themselves look better. Yet many of our rewards are for respect and holding our reputation which if it wasn’t for our bond as a group would not matter to others. According to Charles Handy and the Bos et al. article and seen in the graphs shown in class, “Trust needs touch” and I think this pertains well to our fraternity as experiences together make our reputation system strong and work well where it would not in other communities.
We give certain points out for attending/creating professional events, social events, community service events, as well as for helping brothers in need during certain difficult times. Basically, our desirable behaviors is for every member of the fraternity to be as actively involved with the fraternity as possible through both building cognitive and emotional trust. And for our fraternity to become closer and stronger as a result of this build in trust and reputation.
Points are given out at every chapter with various amounts allowed for certain standardized events. Brothers can also suggest giving points to other brothers for certain good deeds either to another brother or the fraternity as a whole. The individual explains why they deserve to be rewarded and the brother being rewarded explains in further detail what he/she did. Then the entire fraternity votes on whether this is to be awarded or not.
Brothers are then rewarded at different times throughout the semester. Brother recognition occurs weekly at chapter and at our banquet each semester. These have only value intrinsically yet reputation and praise from the whole brotherhood is valued by all of us. For more concrete rewards “familys” within our fraternity with the highest total points each semester get the first picks for “littles” who are new fraternity members who will join that family. Family pride is also a driving force in working towards this goal. As Resnick, et. al. explain in “Reputation Systems,” a good design will cause users to expect past interactions to influence future interactions. Past interactions with others within our fraternity builds a reputation and trust within the group allowing for others to understand and predict their future behavior.
Our scheme could be manipulated if our ties as a whole were not strong. One could collude to give each other points as to make themselves look better. Yet many of our rewards are for respect and holding our reputation which if it wasn’t for our bond as a group would not matter to others. According to Charles Handy and the Bos et al. article and seen in the graphs shown in class, “Trust needs touch” and I think this pertains well to our fraternity as experiences together make our reputation system strong and work well where it would not in other communities.
Monday, March 30, 2009
Assignment 7: Reputation @ CIT [will gunn]
At the beginning of the semester, I started working at the CIT HelpDesk as a consultant. There are a handful of us who work together, and trust is very important. We need to trust each other, and customers need to trust us. We rely on each other's expertise when it comes to troubleshooting, so we need to know that the person we asks is knowledgeable. Even more important, the users who call us need to know that we can be trusted. They are often novice computer users, and they rely solely on what we tell them. They need to know that we are competent and can solve their issues with minimal headache. The current system we use amongst ourselves is very informal. We have mostly learned who knows what, and we have built up a trust through repeated interactions with each other.
There are several desired behaviors that could be used as inputs for a reputation system. We want all of us to be accurate (obviously). We need to know how to solve the problem. We also need to be good communicators, as sometimes the users who call are far from experts. Consistency is also very important. Our work will be much more efficient if we give uniform answers and each consultant is as well-trained as the others.
These behaviors can be (and are) overseen by our supervisors. They pay attention and note how successful we are. They have various ways to measure who is sharp, and who isn't performing as well as they could be. Also, some callers are asked to give feedback and rate how effective the consultant they talked to was. If we perform well, we can receive a raise or a promotion. Also, not as tangibly, the consultants with more expertise definitely earn the respect of the rest of us, and we are willing to defer to them because they are trusted.
The system could be beaten, but it would take a lot of effort. For example, we could get friends to call up with easy to fix problems, and it would seem like we solved every case we encountered. We could appear smarter by to callers by giving an overly complex solution to their problem...but I doubt that would make them as happy as a simple fix would. For the most part though, we learn from each other and help each other out. We're more concerned with appearing reputable to callers, and there is little incentive to "cheat" to better yourself, especially since it requires more effort to game the system than to just do good work. By doing a quality job helping users, the supervisors will notice and your reputation will rise.
There are several desired behaviors that could be used as inputs for a reputation system. We want all of us to be accurate (obviously). We need to know how to solve the problem. We also need to be good communicators, as sometimes the users who call are far from experts. Consistency is also very important. Our work will be much more efficient if we give uniform answers and each consultant is as well-trained as the others.
These behaviors can be (and are) overseen by our supervisors. They pay attention and note how successful we are. They have various ways to measure who is sharp, and who isn't performing as well as they could be. Also, some callers are asked to give feedback and rate how effective the consultant they talked to was. If we perform well, we can receive a raise or a promotion. Also, not as tangibly, the consultants with more expertise definitely earn the respect of the rest of us, and we are willing to defer to them because they are trusted.
The system could be beaten, but it would take a lot of effort. For example, we could get friends to call up with easy to fix problems, and it would seem like we solved every case we encountered. We could appear smarter by to callers by giving an overly complex solution to their problem...but I doubt that would make them as happy as a simple fix would. For the most part though, we learn from each other and help each other out. We're more concerned with appearing reputable to callers, and there is little incentive to "cheat" to better yourself, especially since it requires more effort to game the system than to just do good work. By doing a quality job helping users, the supervisors will notice and your reputation will rise.
Assignment 7: Rank & Reputation [Daniel Gustafson, dgg67]
For this assignment I would like to once again discuss the Taekwondo martial arts school where I trained in high school. While I trained, I began to teach as well, and by the time I left the school, I was a 1st degree black belt and one of four primary instructors at the school. Our programs were mostly oriented towards young children, necessitating each one of us to develop a relationship with the parents of our students. Few parents will trust their children to any-old after school program, so we relied on their trust and our reputation for the success of our business. There were also many students (both children and adults) who would assist with class instruction, so it became very important to display the responsibility levels of instructors as well as students.
The most obvious reputation system in our community is your belt rank/color. There are nine levels of color-belts, the order of which was well known among our students. Belt rank serves essentially as an implicit reputation system. The higher your rank, the more trusted you were to provide correct advice, training tips, or technique demonstrations. Students were awarded with new ranks at testings upon demonstrating their martial arts abilities, their responsibility, and their dedication as a way to recognize their training and elevated level of trust the school was placing in them. Our school had many high-ranking students, so it was also important to designate who the official instructors were. A student-instructor wears a red-white-blue collar, a trainee instructor wears a red collar, a level 2 trainee instructor wears a red/black collar, a lvl 3 wears a black/red/black collar, and a fully-certified instructor wears a full black collar. In a martial arts school, easily visible reputation systems such as these are necessary to quickly command attention and respect from the students. Additionally, parents could easily tell who the primary instructors were, putting them much more at ease about leaving care of their children in our care.
Our school benefited from a very strong reputation. Our school been serving the same community, in the same building for over fifteen years. Additionally, our school was part of a much larger organization of martial arts schools. All schools within this organization had identical rank and reputation systems as described above, so at larger multi-school functions (such as tournaments, testings, or training camps) students and parents were always familiar with the reputation system and could know who to trust. While it is possible to easily fake the conventional signals displayed on one's uniform, behavior such as this within a martial arts organization is nearly unacceptable, non-existent, as well as relatively fruitless. Without the aggregate knowledge described by one's uniform, the display signals are relatively useless, and it would not be difficult to “uncover” such deception (as knowledge is the ultimate test of your uniform's display signals.
The most obvious reputation system in our community is your belt rank/color. There are nine levels of color-belts, the order of which was well known among our students. Belt rank serves essentially as an implicit reputation system. The higher your rank, the more trusted you were to provide correct advice, training tips, or technique demonstrations. Students were awarded with new ranks at testings upon demonstrating their martial arts abilities, their responsibility, and their dedication as a way to recognize their training and elevated level of trust the school was placing in them. Our school had many high-ranking students, so it was also important to designate who the official instructors were. A student-instructor wears a red-white-blue collar, a trainee instructor wears a red collar, a level 2 trainee instructor wears a red/black collar, a lvl 3 wears a black/red/black collar, and a fully-certified instructor wears a full black collar. In a martial arts school, easily visible reputation systems such as these are necessary to quickly command attention and respect from the students. Additionally, parents could easily tell who the primary instructors were, putting them much more at ease about leaving care of their children in our care.
Our school benefited from a very strong reputation. Our school been serving the same community, in the same building for over fifteen years. Additionally, our school was part of a much larger organization of martial arts schools. All schools within this organization had identical rank and reputation systems as described above, so at larger multi-school functions (such as tournaments, testings, or training camps) students and parents were always familiar with the reputation system and could know who to trust. While it is possible to easily fake the conventional signals displayed on one's uniform, behavior such as this within a martial arts organization is nearly unacceptable, non-existent, as well as relatively fruitless. Without the aggregate knowledge described by one's uniform, the display signals are relatively useless, and it would not be difficult to “uncover” such deception (as knowledge is the ultimate test of your uniform's display signals.
Assignment 7 - Kayla Fang (kkf6)
Living in an apartment with three other who are also my friends, sharing amenities, utilities, rent, and space, creates a community based on a lot of trust. Many responsibilities are important to the physical and emotional well-being of our apartment like buying toilet paper, cleaning up the kitchen, taking out the trash/recycling, and planning events in our shared physical and acoustic spaces.
We want a reputation system to keep track of how much we contribute to maintaining our environment and also how much we degrade our environment with negative behaviors, such as being loud, messy, and free-riding. Trust matters a lot because we do not keep track of each others’ behaviors—for example, the huge mess in the middle of the living room could be created by anybody. We trust each other to pay rent, contribute to the electricity bill, and take care of each others’ possessions that are in the common space.
Naturally, many people are inherently lazy so sometimes it can be hard to make sure everybody does their part. One thing that every single person in my apartment likes is Sour Patch Kids candy. A possible rewards system could involve having a large supply of Sour Patch Kids and every person can have a small cylinder container (like a thermometer) next to the door with her name on it—whenever we find out about something constructive and positive they do, we can add Sour Patch Kids to the bag, and if we are unhappy with their mess, we can take out Sour Patch Kids. This way, we can each gauge how satisfied our roommates in general are with our participation. Every week we can get together (on Sundays or something) and eat all of the Sour Patch Kids in our containers and start afresh. This way, we also have a regular meeting in which we can acknowledge each other’s contributions and bloopers. Resnick discusses the challenges of eliciting, distributing, and aggregating feedback in his paper on Reputation Systems, all of which the SPK system attempts to address: the “fun” aspect of candy should encourage my roommates to participate regularly,; distribution is easy because our feedback is in a container next to our front door; lastly, aggregation is simply the total volume of Sour Patch Kids in the container as an aggregation of everybody’s feedback.
This system can be manipulated by each person by adding extra Sour Patch Kids to her own container to make it look like the general approval is higher than it is so that the other roommates will think that she is doing what she is supposed to.
We want a reputation system to keep track of how much we contribute to maintaining our environment and also how much we degrade our environment with negative behaviors, such as being loud, messy, and free-riding. Trust matters a lot because we do not keep track of each others’ behaviors—for example, the huge mess in the middle of the living room could be created by anybody. We trust each other to pay rent, contribute to the electricity bill, and take care of each others’ possessions that are in the common space.
Naturally, many people are inherently lazy so sometimes it can be hard to make sure everybody does their part. One thing that every single person in my apartment likes is Sour Patch Kids candy. A possible rewards system could involve having a large supply of Sour Patch Kids and every person can have a small cylinder container (like a thermometer) next to the door with her name on it—whenever we find out about something constructive and positive they do, we can add Sour Patch Kids to the bag, and if we are unhappy with their mess, we can take out Sour Patch Kids. This way, we can each gauge how satisfied our roommates in general are with our participation. Every week we can get together (on Sundays or something) and eat all of the Sour Patch Kids in our containers and start afresh. This way, we also have a regular meeting in which we can acknowledge each other’s contributions and bloopers. Resnick discusses the challenges of eliciting, distributing, and aggregating feedback in his paper on Reputation Systems, all of which the SPK system attempts to address: the “fun” aspect of candy should encourage my roommates to participate regularly,; distribution is easy because our feedback is in a container next to our front door; lastly, aggregation is simply the total volume of Sour Patch Kids in the container as an aggregation of everybody’s feedback.
This system can be manipulated by each person by adding extra Sour Patch Kids to her own container to make it look like the general approval is higher than it is so that the other roommates will think that she is doing what she is supposed to.
Assignment #7: Daniela Retelny
A community that I am part of is the Information Science Student Association, which is a club dedicated to bringing together Information Science students, keeping people updated about the major as well as helping members network with others in the field. A couple of weeks ago I was elected to be the Student Curriculum Representative. This job entails meeting twice a month with faculty to discuss the major requirements and make improvements.
Trust plays a big role in the Student Curriculum Representative position. By being selected for the position, students trust that I will get their opinions and messages across to the Information Science faculty. However, I also need to trust the ISSA members and their messages, because without their input I do not have much information to share at the curriculum meetings.
The type of trust involved in the ISSA community is cognitive trust. Trust is built between individuals through interaction, experience and longevity. Therefore, a recommender system would need to take these things into account. Desirable behaviors include students sharing nonbiased, accurate and true statements about what they think needs to be changed in the major. Because the group communicates primarily face-to-face at our meetings and follows up via email, the group relies on cooperative behavior. Our face-to-face meetings relate to Handy and Bos et. al’s article on “Trust Needs Touch.”
A potential recommender system for this community, particularly for my job as Student Curriculum Representative would be an online forum where all members of the ISSA have a profile and can start new threads for certain topics. Group members could then rate each other’s forums topics and ideas as well as respond to each other’s posts. There would also be a vote feature, where students can say whether or not they agree with their group members posts. By knowing that their names are exposed and that their posts are going to be rated by other Information Science students, people will have more of an incentive to keep their posts accurate and beneficial. The authors of the posts with the most votes, as well as those who have been active members for a certain period of time would get a higher status. These “experts” could act as resources to incoming and current students and maybe even gain extra “brownie” points by faculty members.
Ultimately I would reward the members of the ISSA by getting their voices heard. Ideally the reward to all of the Information Sciences majors is a major filled with thoughtful, useful classes that will prepare them well for their future endeavors. However, on the site, students would be rewarded by gaining points depending on how much other members agree with their post. These points would gain them status and strengthen their reputation.
Trust plays a big role in the Student Curriculum Representative position. By being selected for the position, students trust that I will get their opinions and messages across to the Information Science faculty. However, I also need to trust the ISSA members and their messages, because without their input I do not have much information to share at the curriculum meetings.
The type of trust involved in the ISSA community is cognitive trust. Trust is built between individuals through interaction, experience and longevity. Therefore, a recommender system would need to take these things into account. Desirable behaviors include students sharing nonbiased, accurate and true statements about what they think needs to be changed in the major. Because the group communicates primarily face-to-face at our meetings and follows up via email, the group relies on cooperative behavior. Our face-to-face meetings relate to Handy and Bos et. al’s article on “Trust Needs Touch.”
A potential recommender system for this community, particularly for my job as Student Curriculum Representative would be an online forum where all members of the ISSA have a profile and can start new threads for certain topics. Group members could then rate each other’s forums topics and ideas as well as respond to each other’s posts. There would also be a vote feature, where students can say whether or not they agree with their group members posts. By knowing that their names are exposed and that their posts are going to be rated by other Information Science students, people will have more of an incentive to keep their posts accurate and beneficial. The authors of the posts with the most votes, as well as those who have been active members for a certain period of time would get a higher status. These “experts” could act as resources to incoming and current students and maybe even gain extra “brownie” points by faculty members.
Ultimately I would reward the members of the ISSA by getting their voices heard. Ideally the reward to all of the Information Sciences majors is a major filled with thoughtful, useful classes that will prepare them well for their future endeavors. However, on the site, students would be rewarded by gaining points depending on how much other members agree with their post. These points would gain them status and strengthen their reputation.
Assignment 7: Emily Wagner
We learned in class that “reputation systems attempt to mediate and automate reputation: to take note of a community-member’s actions; assess the community’s reaction to them; and keep a running tally of the history of these actions.” A community that I am a part of is my sorority, and we have recently instated a points system that acts as a reputation system. At first it was not very effective because no one was really sure what the rules and requirements were and we learned that more specific sets of rules have greater influence on behavior. Now the points system has very clear guidelines and seems to be very efficient.
Trust is required in this community because without it the chapter would not be united and would not be able to run at all. Trust is required amongst friends as well as leaders and as a whole we need to trust that the officers are doing everything fairly and that no one is cheating the system. The desired behaviors are mainly going to events and activities. There are certain points that you must achieve in order to “break even”, ie. Points are deducted if girls do not go to certain events.
In this system, we can see which girls are not going to mandatory meetings, philanthropy events, or other activities around campus. However, girls are also rewarded extra points for going to activities that are not required, going to other house’s philanthropy events, and for helping out officers in ways that are considered going above and beyond their responsibilities. Doing more than is expected of us is truly the desired behavior. This is rewarded not only by verbal recognition every week, but also with gift certificates. The biggest reward is to the lineage with the most points, which receives a limo ride to our formal at the end of the semester. Making this reward a group prize encourages more people to participate as to not let people down in the end.
The main way this could be manipulated is if the girl in charge of the points system were to unfairly distribute points. However, due to the trust in each other we are assured that this does not happen because the officer is elected to her position due to the fact that we trust her. We trust that she has fair judgment in deciding if excuses are legitimate, and we trust that she follows the rules just as we do.
Trust is required in this community because without it the chapter would not be united and would not be able to run at all. Trust is required amongst friends as well as leaders and as a whole we need to trust that the officers are doing everything fairly and that no one is cheating the system. The desired behaviors are mainly going to events and activities. There are certain points that you must achieve in order to “break even”, ie. Points are deducted if girls do not go to certain events.
In this system, we can see which girls are not going to mandatory meetings, philanthropy events, or other activities around campus. However, girls are also rewarded extra points for going to activities that are not required, going to other house’s philanthropy events, and for helping out officers in ways that are considered going above and beyond their responsibilities. Doing more than is expected of us is truly the desired behavior. This is rewarded not only by verbal recognition every week, but also with gift certificates. The biggest reward is to the lineage with the most points, which receives a limo ride to our formal at the end of the semester. Making this reward a group prize encourages more people to participate as to not let people down in the end.
The main way this could be manipulated is if the girl in charge of the points system were to unfairly distribute points. However, due to the trust in each other we are assured that this does not happen because the officer is elected to her position due to the fact that we trust her. We trust that she has fair judgment in deciding if excuses are legitimate, and we trust that she follows the rules just as we do.
Assignment #7 - Angel M. Villegas
A community I am involved with plays paintball and I think it would be great to have a reputation system for this community. Actually it would be useful to incorporate many rating systems into this reputation scheme. Since there are many facets to playing paintball, I am sure that my thoughts for ranking may not be inclusive of all possible scales. I believe a reputation system will help players form teams with others who they have never played with. This way the team might not know the individual personally but will have a clue for what their skill set is, if they are trust worthy, and how that can be incorporated into the teams mechanics. Trust is important because some players have different motives for playing the game, some are really competitive, others play just for fun, and then some play to screw up the game and the "rules" in place.
For example accuracy should be a skill rated within the system. This can be measured in a number of ways. If playing in a team setting then special paintballs can be created that are shades of the teams color, thus distinguishing an individual’s performance by tallying up the number of paintballs on the opposing team and the number missing. Another example would be to set up a ranking for a player shooting their own teammates. This would be a big factor of whether the team could trust you or not.
When playing paintball a team can decide to allow the player to join the team or reject them according to their reputation. Not all ratings in the system would be relevant to this scenario but knowing if a person generally is a trader to the team can have some bearing on their willingness to trust the person to cover their back. Having a good or at least a reputation for playing the game fair is rewarded by being inducted into a team.
This system would require that all the users are a part of a system that is able to record them to the details needed to uniquely identify them. If the system in place could not uniquely identify a person then anyone could set up a bogus profile and get rated one way and truly be the type of player to turn on his teammates. Also if the team works together to single out someone they can cause the person to lower their reputation if they stand in front of them shooting or take his gun and cause the players reputation to lower due to their foolish conduct.
For example accuracy should be a skill rated within the system. This can be measured in a number of ways. If playing in a team setting then special paintballs can be created that are shades of the teams color, thus distinguishing an individual’s performance by tallying up the number of paintballs on the opposing team and the number missing. Another example would be to set up a ranking for a player shooting their own teammates. This would be a big factor of whether the team could trust you or not.
When playing paintball a team can decide to allow the player to join the team or reject them according to their reputation. Not all ratings in the system would be relevant to this scenario but knowing if a person generally is a trader to the team can have some bearing on their willingness to trust the person to cover their back. Having a good or at least a reputation for playing the game fair is rewarded by being inducted into a team.
This system would require that all the users are a part of a system that is able to record them to the details needed to uniquely identify them. If the system in place could not uniquely identify a person then anyone could set up a bogus profile and get rated one way and truly be the type of player to turn on his teammates. Also if the team works together to single out someone they can cause the person to lower their reputation if they stand in front of them shooting or take his gun and cause the players reputation to lower due to their foolish conduct.
Assignment 7 - Jeffrey Hertzberg
I've been a member of small online art and writing sharing community for long time. It's a community where members can come and share pieces they've been writing on and also get critiques and constructive criticism from other members.
Trust is built over time in the community along with reputation. As people post more of their work and critiques, as well as just casual banter amongst the forums they gain the trust of other members. First, there's just to make sure that what the person is posting is actually their own work and not stolen from someone else. Also, it can be perceived through the way that they speak and carry out their comments whether or not they are experienced and bring a decent amount of prior knowledge in the field they are most involved in.
The reputation system I'd put within the community could be viewed from multiple standpoints. First there are competitions/contests within the various forums with specific themes in creating either a work of art or writing, that may also specify the style that must be used. Each user can vote on the submissions and the winner will gain titles or small awards. There are also a set of "Muses" that run each forum and make sure their competitions and events go smoothly that have a basically constant higher tier reputation than the rest. Finally, merely the amount of comments and compliments a member receives when they submit a new work helps to establish their reputation as either a better or worse creator with the other members of the site.
The desirable behaviors would be to create amazing pieces of art/writing that inspire others to comment, critique, and create more of their own. Also, to create pieces that will be approved of by enough people that you will win the contest for that week. Rewards of course I've been over already with the receive of forum titles, and signature rewards. A signature is posted with every post you make and so, therefore, would your reward be as well.
Hopefully not much of the scheme could be manipulated. However, members might try to rig contests, or abuse the rights of posting awards within their sigs. There is also of course the worry that people may be using other's copyrighted works to construct their own pieces and not adding enough to really be thought of as original. Then we have to reevaluate the level of trust we place in people, and of course their reputation.
Trust is built over time in the community along with reputation. As people post more of their work and critiques, as well as just casual banter amongst the forums they gain the trust of other members. First, there's just to make sure that what the person is posting is actually their own work and not stolen from someone else. Also, it can be perceived through the way that they speak and carry out their comments whether or not they are experienced and bring a decent amount of prior knowledge in the field they are most involved in.
The reputation system I'd put within the community could be viewed from multiple standpoints. First there are competitions/contests within the various forums with specific themes in creating either a work of art or writing, that may also specify the style that must be used. Each user can vote on the submissions and the winner will gain titles or small awards. There are also a set of "Muses" that run each forum and make sure their competitions and events go smoothly that have a basically constant higher tier reputation than the rest. Finally, merely the amount of comments and compliments a member receives when they submit a new work helps to establish their reputation as either a better or worse creator with the other members of the site.
The desirable behaviors would be to create amazing pieces of art/writing that inspire others to comment, critique, and create more of their own. Also, to create pieces that will be approved of by enough people that you will win the contest for that week. Rewards of course I've been over already with the receive of forum titles, and signature rewards. A signature is posted with every post you make and so, therefore, would your reward be as well.
Hopefully not much of the scheme could be manipulated. However, members might try to rig contests, or abuse the rights of posting awards within their sigs. There is also of course the worry that people may be using other's copyrighted works to construct their own pieces and not adding enough to really be thought of as original. Then we have to reevaluate the level of trust we place in people, and of course their reputation.
International Trust- Assignment 7, Beth
I am part of a student-run international organization called AIESEC that has over 800 local committees (LCs) at universities worldwide. In order to work towards a mission of cultural understanding, each of the local LCs sets up internships in their area and advertises them in an online forum. Students search through the options looking for what kind of job they want and where they want to live. But, these are not the only things that matter to a student traveling across the world. When a student arrives in a new country, the LC is responsible for taking care of many details of their arrival, including rides from the airport, finding them a place to stay, and getting them a cell phone. This is a major point in the process where trust is crucial. The exchange student needs to be able to trust that there will be people on the other side of the world preparing for their arrival. They are relying on busy students to take care of them.
According to Charles Handy and the Bos et al. article, “Trust needs touch”. Our organization has dozens of international conferences a year for this very reason. By meeting in person and talking about non-work related subjects, we are more able to trust the receiver of our students and set up exchange programs with the LC of someone we know. But, what if we can’t meet someone from every country?
A reputation system would help build trust in our existing online forum. In order to establish a sustained entity for expectations, which Resnick deems necessary, each LC could have a profile to display data about past experiences with exchange students. This information would include: number of internships raised with a higher number showing a more sustained LC, number of exchange students brought in showing experience with exchange students, and a combined rating from all past exchange students based on the comfort and support of the LC in their experience. This feedback would be visible along side the job posting to allow students to judge how much they can trust the LC to cater to their needs when they arrive at their host country.
LCs want to take in exchange students because it increases cultural awareness in the LC and surrounding community…and also brings in some money to the LC. With this motivation, LCs strive to please their exchange students. Future exchange students might trust the whole program more if it had a rating system, as Resnick points out that agreeing to be rated shows “an indication of higher-quality services” (47). Therefore, whether or not LCs manipulate the system by bribing their exchange students to give them higher ratings, the reputation system would help the organization (and human kind) succeed in more exchanges and promote cultural awareness.
According to Charles Handy and the Bos et al. article, “Trust needs touch”. Our organization has dozens of international conferences a year for this very reason. By meeting in person and talking about non-work related subjects, we are more able to trust the receiver of our students and set up exchange programs with the LC of someone we know. But, what if we can’t meet someone from every country?
A reputation system would help build trust in our existing online forum. In order to establish a sustained entity for expectations, which Resnick deems necessary, each LC could have a profile to display data about past experiences with exchange students. This information would include: number of internships raised with a higher number showing a more sustained LC, number of exchange students brought in showing experience with exchange students, and a combined rating from all past exchange students based on the comfort and support of the LC in their experience. This feedback would be visible along side the job posting to allow students to judge how much they can trust the LC to cater to their needs when they arrive at their host country.
LCs want to take in exchange students because it increases cultural awareness in the LC and surrounding community…and also brings in some money to the LC. With this motivation, LCs strive to please their exchange students. Future exchange students might trust the whole program more if it had a rating system, as Resnick points out that agreeing to be rated shows “an indication of higher-quality services” (47). Therefore, whether or not LCs manipulate the system by bribing their exchange students to give them higher ratings, the reputation system would help the organization (and human kind) succeed in more exchanges and promote cultural awareness.
Assignment 7: Joyce Lee (jl579)
There are many sites on the internet that allow for both amateur and professional artists to upload and share their work without having to register their own webspace. Deviantart.com or Conceptart.org are two such sites, of which I am a member. Reputation plays a large part in the system, where getting acknowledge or noticed is something everyone strives for, but at the same time, you do not want to be subject to someone's simple five-minute doodles without effort put into them.
Trust is a large issue for sites like these. User content is uploaded at the artist's discretion, and though there is a moderation team, it is impossible to check every single image uploaded every moment. Consequently, users trust other users to appropriately tag their work if there are warnings on it for gore, nudity, etc. If not, users also trust each other to "flag" an image and notify a moderator of what rules it is breaking. Users, in other words, trust others not to upload inappropriate imagery where we are 'vulnerable' (as according to the definition in class) to seeing it. The most important aspect of this system is an active flow of feedback from users, as Resnick describes.
The end result is a helpful and inspiring resource for artists - a collection of artworks of high caliber that others can critique and draw techniques from. The reward for a good 'reputation' is acknowledgment. More viewers will be attracted to your page if you have not gone around breaking rules. If you consistently produce good artwork, the number of people following your work will increase. For artists, this exposure is highly important in career searches, which many of the members are currently in the process of. This reward system inspires future involvement in hopes of accruing more viewers. Involvement in the community is expected to be long-lived: another element that Resnick stresses.
Optimistically, users will not upload inappropriate images at all, and there would be no flagging. Barring that, users will flag the right kind of images - images that don't follow the rules, are too explicit, and not images that they simply don't like. Ideally, moderators will judge to take down an image objectively, and not play favorites. Unfortunately, users can still get through the moderation team's notice with inappropriate work. If viewers do not care enough about rules to report a violation, then it remains untouched. Repeated offenders can also create new accounts, since that is not limited.
Trust is a large issue for sites like these. User content is uploaded at the artist's discretion, and though there is a moderation team, it is impossible to check every single image uploaded every moment. Consequently, users trust other users to appropriately tag their work if there are warnings on it for gore, nudity, etc. If not, users also trust each other to "flag" an image and notify a moderator of what rules it is breaking. Users, in other words, trust others not to upload inappropriate imagery where we are 'vulnerable' (as according to the definition in class) to seeing it. The most important aspect of this system is an active flow of feedback from users, as Resnick describes.
The end result is a helpful and inspiring resource for artists - a collection of artworks of high caliber that others can critique and draw techniques from. The reward for a good 'reputation' is acknowledgment. More viewers will be attracted to your page if you have not gone around breaking rules. If you consistently produce good artwork, the number of people following your work will increase. For artists, this exposure is highly important in career searches, which many of the members are currently in the process of. This reward system inspires future involvement in hopes of accruing more viewers. Involvement in the community is expected to be long-lived: another element that Resnick stresses.
Optimistically, users will not upload inappropriate images at all, and there would be no flagging. Barring that, users will flag the right kind of images - images that don't follow the rules, are too explicit, and not images that they simply don't like. Ideally, moderators will judge to take down an image objectively, and not play favorites. Unfortunately, users can still get through the moderation team's notice with inappropriate work. If viewers do not care enough about rules to report a violation, then it remains untouched. Repeated offenders can also create new accounts, since that is not limited.
Assignment 7: Eric Dial
An online community that I am a part of footballsfuture.com. This website provides in depth coverage of all current NFL teams, and allows users to post blogs about specific events or happenings around the league, such as the NFL Draft or big trades. Everything, from what teams are talking to what player to who will make the playoffs next year is discussed. I am a HUGE fan of the St. Louis Rams and this is the website I go to for all the latests news about my team and other teams in the league.
There has to be a lot of trust between members of this community. Many posts will be about rumors and stories that have been heard on the television and radio or read on a website, and if it affects your team or the NFL, you have to trust that these sources are credible. If an event, such as a big-name player being traded to another team is occurring, but the person that posted did not list any website or source, then most members know that this is speculation. However, most of the time, when anything like this happens, almost always the user lists a source. Although there is not face to face interaction, I believe trust still exists among members out of the respect we have for the teams and players.
The desirable behaviors for this community are blog posts, signature creations, and overall sharing of information throughout the league. I've already touched on the posts, but another big thing on this website is creating your own signature. Signatures can be anything designed to highlight your favorite team, favorite player, or who you think will be the best draft pick this year. I myself don't have a signature just because I don't think I'm creative enough, but I think some of the ones people have come up with are really creative and cool. Rewards for this sight are minimal. The most you'll get is someone commenting on your post that it was a good story or interesting article. Others will tell you that your signature is cool and it was a good design.
A good reputation system for this website would be a rating of the user's posts and their sources of information. If a user has a really interesting story about a player and a source to back it up, they would get a higher rating from other users than would a less interesting story with no source. This rating would be made public, as it would appear next to the users screen name, thus revealing whether it be worthwhile for another user to read their post.
There has to be a lot of trust between members of this community. Many posts will be about rumors and stories that have been heard on the television and radio or read on a website, and if it affects your team or the NFL, you have to trust that these sources are credible. If an event, such as a big-name player being traded to another team is occurring, but the person that posted did not list any website or source, then most members know that this is speculation. However, most of the time, when anything like this happens, almost always the user lists a source. Although there is not face to face interaction, I believe trust still exists among members out of the respect we have for the teams and players.
The desirable behaviors for this community are blog posts, signature creations, and overall sharing of information throughout the league. I've already touched on the posts, but another big thing on this website is creating your own signature. Signatures can be anything designed to highlight your favorite team, favorite player, or who you think will be the best draft pick this year. I myself don't have a signature just because I don't think I'm creative enough, but I think some of the ones people have come up with are really creative and cool. Rewards for this sight are minimal. The most you'll get is someone commenting on your post that it was a good story or interesting article. Others will tell you that your signature is cool and it was a good design.
A good reputation system for this website would be a rating of the user's posts and their sources of information. If a user has a really interesting story about a player and a source to back it up, they would get a higher rating from other users than would a less interesting story with no source. This rating would be made public, as it would appear next to the users screen name, thus revealing whether it be worthwhile for another user to read their post.
Assignment 7 (hrs34)
One small scale community that I am a part of is my house. I live in college town with 6 others and we have a house to ourselves. In an effort to make close quarter livings as painless as possible, we have already instituted a sort of reputation system.
A reputation system, in the context that we will discus it, is a way of measuring one’s perceived trustworthiness and communal status by others. The example used by Resnick et al, was the rating system implemented by eBay (-1, 0, 1). The purpose of their system is to provide incentive to perform truthful and professional interactions. If a seller is dishonest or slow, they will receive a bad rating and will have a hard time selling in the future.
The reputation system used in our house is less of a rating system and more of an accountability system. We created a chore wheel with the 7 house members and 7 weekly chores. To further it as a reputation system, we could also create a chart so that when the chore is complete, you could check off that it was done. As is, a chore could be forgotten about in a given week and may go unnoticed. This is why trust is important in our community. We are all trusting that the others in the house will accomplish their task by the end of the week. Chores include things like taking out the garbage, and if this is not done we’ll have a stinky mess on our hands.
We trust that it will get done. We desire that it will get done early in the week, or whenever necessary. Rewards for doing so are present, but are currently intangible. By that, I mean that if someone does a good job with chores, they may get a high five, but there are no physical rewards such as an easier chore the next week.
As mentioned, creating a chart would further our reputation system because it contributes the three necessary elements described by Resnick:
• Long-lived entities that inspire an expectation of future interaction;
- a chart keeps permanent records.
• Capture and distribution of feedback about current interactions (such information must be visible in the future)
- easy to view when others have completed tasks
• Use of feedback to guide trust decisions.
-if a check is missing, trust is implicitly lowered.
A reputation system, in the context that we will discus it, is a way of measuring one’s perceived trustworthiness and communal status by others. The example used by Resnick et al, was the rating system implemented by eBay (-1, 0, 1). The purpose of their system is to provide incentive to perform truthful and professional interactions. If a seller is dishonest or slow, they will receive a bad rating and will have a hard time selling in the future.
The reputation system used in our house is less of a rating system and more of an accountability system. We created a chore wheel with the 7 house members and 7 weekly chores. To further it as a reputation system, we could also create a chart so that when the chore is complete, you could check off that it was done. As is, a chore could be forgotten about in a given week and may go unnoticed. This is why trust is important in our community. We are all trusting that the others in the house will accomplish their task by the end of the week. Chores include things like taking out the garbage, and if this is not done we’ll have a stinky mess on our hands.
We trust that it will get done. We desire that it will get done early in the week, or whenever necessary. Rewards for doing so are present, but are currently intangible. By that, I mean that if someone does a good job with chores, they may get a high five, but there are no physical rewards such as an easier chore the next week.
As mentioned, creating a chart would further our reputation system because it contributes the three necessary elements described by Resnick:
• Long-lived entities that inspire an expectation of future interaction;
- a chart keeps permanent records.
• Capture and distribution of feedback about current interactions (such information must be visible in the future)
- easy to view when others have completed tasks
• Use of feedback to guide trust decisions.
-if a check is missing, trust is implicitly lowered.
Reputation System on the Team
The reputation system on the women's squash team has a variety of measures and contributing factors. The most obvious reputation system is the ladder. The ladder, numbered 1-12, is almost purely based on the player’s ability on the court. This is a combination of her racquet skills, speed, consistency, and mental toughness. It is however, occasionally altered (not manipulated) by an individuals commitment to the team, presence at practice, and input at practice (how hard the player is working). This reputation system could be manipulated by the coach's biased opinion, as she has the ultimate say the in the playing ladder. This manipulation is often seen, not to reward or punish a player on the team; however, to strategically match our player with a specific member on the opposing team, a move commonly referred to as "stacking". For example, if Cornell #1 player has a specific weakness against Harvard's #1 player, and the coach thinks that Cornell #2 player would have a better chance of beating Harvard's #1, then the coach might switch Cornell's #1 and 2 players, or "stack" the ladder. This is not allowed in the College Squash Association (CSA), however many coaches manipulate the ladder in subtle enough ways to get away with it.
Another reputation system, which is less structured than a numbered ladder, is our awards system at the end of the season. As ours is coming up, each player on the team has to think about which player she wants to nominate for the following awards: Sportsmanship Award, Most Improved Player, Best Match of the Season, Most Valuable Player, and Captain. Each award rewards different admirable behavior of the specific athlete. The kind of behavior is indicated by the name of the award- (I won't bore you stating the obvious). Captains however, are generally elected if the individual exemplifies dedication to the team, leadership and initiative, enthusiasm, and excellence on the court. This reputation system works based on nominations from each individual player. You are not allowed to vote for yourself. The system could be manipulated by individuals voting based on personal ties rather than the appropriate characteristics; In other words, girls voting for their friends.
Although immeasurable, the underlying reputation system on this team is respect for each other, which is where and why trust is so important. We trust each other to be committed to the team—to show up physically and mentally for practice, to motivate each other, and to never give up. I trust that my teammates are giving 100% to the team, as they trust that I am doing the same.
Another reputation system, which is less structured than a numbered ladder, is our awards system at the end of the season. As ours is coming up, each player on the team has to think about which player she wants to nominate for the following awards: Sportsmanship Award, Most Improved Player, Best Match of the Season, Most Valuable Player, and Captain. Each award rewards different admirable behavior of the specific athlete. The kind of behavior is indicated by the name of the award- (I won't bore you stating the obvious). Captains however, are generally elected if the individual exemplifies dedication to the team, leadership and initiative, enthusiasm, and excellence on the court. This reputation system works based on nominations from each individual player. You are not allowed to vote for yourself. The system could be manipulated by individuals voting based on personal ties rather than the appropriate characteristics; In other words, girls voting for their friends.
Although immeasurable, the underlying reputation system on this team is respect for each other, which is where and why trust is so important. We trust each other to be committed to the team—to show up physically and mentally for practice, to motivate each other, and to never give up. I trust that my teammates are giving 100% to the team, as they trust that I am doing the same.
Assignment 7
The online community I am choosing to analyze is that of xanga. Xanga is an online blogging site that allows one to post thoughts and feelings, design a personal interface, make friends, and join groups as well as a variety of other tasks.
Bos et al. defines trust as “willingness to be vulnerable, based on positive expectations about the actions of others”. This idea can be shown in the xanga community in that new users are more likely to be trustworthy right away when faced with the successful journaling done by older members. The newer members are less hesitant in revealing secrets and private thoughts due to the existence and experience of the community. Trust matters in this community because you are essentially using the site as a journal. Virtual strangers may be reading your innermost thoughts and desired. One may think that this would cause many to shy away from making their posts public, but the effect is quite the opposite. Many bloggers feel safe in the xanga community.
Xanga is a reputation system since one can see how many subscribers, comments and recommended posts a blogger has and use that information to judge him/her. Also, as Resnick states, "an expectation that people will consider one another’s pasts in future interactions constrains behavior in the present." This statement is probably the driving force behind why people leave mostly nice and positive comments.
Desirable behaviors in this community include commenting, recommending, designing interfaces and subscribing to blogs. Commenting is usually rewarded with commenting or recommending or subscribing. For example, if person A comments on person B’s blog, then person B will comment back on person A’s blog. However, sometimes person B might go further and subscribe to person A’s blog and vice versa. There is somewhat of a hierarchy in rewarding. Comments are the easiest reward and desirable behavior. Next are recommendations, then subscribing to blogs. The design of interfaces is a behavior that can be custom made to a person’s personality (and thereby rewarded more directly) or can be left to the public by publishing codes for everyone to take.
Of course the main desirable behavior is to have interesting and appealing blog posts. And this behavior is rewarded by all which was mentioned above.
This scheme is often manipulated since people will comment haphazardly across blogs hoping to get others to comment back. Those who design interfaces often plead and beg people to subscribe, offering free personalized interfaces as an incentive. Even though this reputation system already had a scheme in plays that rewards desirable behaviors, it can and often is manipulated.
Bos et al. defines trust as “willingness to be vulnerable, based on positive expectations about the actions of others”. This idea can be shown in the xanga community in that new users are more likely to be trustworthy right away when faced with the successful journaling done by older members. The newer members are less hesitant in revealing secrets and private thoughts due to the existence and experience of the community. Trust matters in this community because you are essentially using the site as a journal. Virtual strangers may be reading your innermost thoughts and desired. One may think that this would cause many to shy away from making their posts public, but the effect is quite the opposite. Many bloggers feel safe in the xanga community.
Xanga is a reputation system since one can see how many subscribers, comments and recommended posts a blogger has and use that information to judge him/her. Also, as Resnick states, "an expectation that people will consider one another’s pasts in future interactions constrains behavior in the present." This statement is probably the driving force behind why people leave mostly nice and positive comments.
Desirable behaviors in this community include commenting, recommending, designing interfaces and subscribing to blogs. Commenting is usually rewarded with commenting or recommending or subscribing. For example, if person A comments on person B’s blog, then person B will comment back on person A’s blog. However, sometimes person B might go further and subscribe to person A’s blog and vice versa. There is somewhat of a hierarchy in rewarding. Comments are the easiest reward and desirable behavior. Next are recommendations, then subscribing to blogs. The design of interfaces is a behavior that can be custom made to a person’s personality (and thereby rewarded more directly) or can be left to the public by publishing codes for everyone to take.
Of course the main desirable behavior is to have interesting and appealing blog posts. And this behavior is rewarded by all which was mentioned above.
This scheme is often manipulated since people will comment haphazardly across blogs hoping to get others to comment back. Those who design interfaces often plead and beg people to subscribe, offering free personalized interfaces as an incentive. Even though this reputation system already had a scheme in plays that rewards desirable behaviors, it can and often is manipulated.
Assignment #7 - Eugene Chang
My house is the community that I probably spend my most time in. My roommates are some of my best friends and it comes out in how we deal with living each other. Trust is important between roommates because we share everything. From food to living in the same space, trusting that your roommates will pay bills fairly, clean up after themselves, do chores, etc. is paramount to a fun and livable situation.
So consider a kind of roommate point reputation system. We’ll have a system that keeps track of how many times a person does a good roommate task. This could range from small tasks like taking out the trash to larger tasks (i.e. grosser) like cleaning the bathroom. Different tasks could have different point values associated with each. This makes it useful to keep track of what a roommate is doing and to see if he is “pulling his weight” for the rest of the house. The reward of doing this scheme is making sure tasks are not unfairly distributed across people. One could also imagine this being a good way to reward someone who does a lot for the house by allowing them to not have to do jobs they wouldn’t want to do.
There’s a few ways this system could be manipulated. People could abuse different jobs by doing them more often than need be. Taking out the trash only really needs to happen every so often. Also other jobs may never get done and there’s no motivation to actually do other jobs. Also, depending on the implementation, there’s a trust issue on even recording what tasks were done and to what extent they were done. For example, perhaps I only swept half the kitchen floor but I mark that I did the entire task. There’s no measurement of quality on completing the task.
So consider a kind of roommate point reputation system. We’ll have a system that keeps track of how many times a person does a good roommate task. This could range from small tasks like taking out the trash to larger tasks (i.e. grosser) like cleaning the bathroom. Different tasks could have different point values associated with each. This makes it useful to keep track of what a roommate is doing and to see if he is “pulling his weight” for the rest of the house. The reward of doing this scheme is making sure tasks are not unfairly distributed across people. One could also imagine this being a good way to reward someone who does a lot for the house by allowing them to not have to do jobs they wouldn’t want to do.
There’s a few ways this system could be manipulated. People could abuse different jobs by doing them more often than need be. Taking out the trash only really needs to happen every so often. Also other jobs may never get done and there’s no motivation to actually do other jobs. Also, depending on the implementation, there’s a trust issue on even recording what tasks were done and to what extent they were done. For example, perhaps I only swept half the kitchen floor but I mark that I did the entire task. There’s no measurement of quality on completing the task.
Sunday, March 29, 2009
Assignment 7: Will Hui
I’m on the mailing list for a software program called Enso. Subscribers trust others not to abuse the list. Specifically, they trust others not to spam, post inappropriate material, or attempt phishing scams. People on the list also trust others to be well-intentioned (i.e. they do not wish to actively harm the community). Poisonous people could drain user enthusiasm and/or suck up time and resources from Enso developers by igniting flame wars and expressing consistent pessimism.
It’s desirable to have members post messages that are on-topic, relevant to most members on the list, friendly (to both the project and its members), and not trivially answered elsewhere. Consider a reward system where the community can rate the quality of each post, perhaps on a scale of 1 (worst) to 5 (best). Users rate posts by emailing the listserv with a rating for a particular post. Posts generating more responses (and responses from different people) would also get a higher quality ranking. These explicit and implicit feedback mechanisms would be combined in some way to generate an overall quality ranking that affects the number of points assigned to that post. A user’s reputation is the weighted average of all his posts, calculated in a manner such that newer posts are given more weight than older posts.
As Resnick, et. al. explain in “Reputation Systems,” a good design will cause users to expect past interactions to influence future interactions. The reward system will create this “shadow of the future” effect by attaching each person’s reputation to every post he makes. The quality of posts made in the past will affect his current rating, which may impact how community members respond to him in the future.
Although we’re using explicit feedback in the form of ratings, a single user shouldn’t be able to affect rankings too drastically. This is because you only get to rate each post once, and the explicit ratings are merged with implicit feedback (responses generated by the post) in order to arrive at the final point value. But if someone subscribes many different email addresses to the list, he could game the system by either using all those accounts to rate a single post down or respond to a given post many times, boosting its score.
It’s desirable to have members post messages that are on-topic, relevant to most members on the list, friendly (to both the project and its members), and not trivially answered elsewhere. Consider a reward system where the community can rate the quality of each post, perhaps on a scale of 1 (worst) to 5 (best). Users rate posts by emailing the listserv with a rating for a particular post. Posts generating more responses (and responses from different people) would also get a higher quality ranking. These explicit and implicit feedback mechanisms would be combined in some way to generate an overall quality ranking that affects the number of points assigned to that post. A user’s reputation is the weighted average of all his posts, calculated in a manner such that newer posts are given more weight than older posts.
As Resnick, et. al. explain in “Reputation Systems,” a good design will cause users to expect past interactions to influence future interactions. The reward system will create this “shadow of the future” effect by attaching each person’s reputation to every post he makes. The quality of posts made in the past will affect his current rating, which may impact how community members respond to him in the future.
Although we’re using explicit feedback in the form of ratings, a single user shouldn’t be able to affect rankings too drastically. This is because you only get to rate each post once, and the explicit ratings are merged with implicit feedback (responses generated by the post) in order to arrive at the final point value. But if someone subscribes many different email addresses to the list, he could game the system by either using all those accounts to rate a single post down or respond to a given post many times, boosting its score.
Assignment 7 -- Adam Towne
I spent three summers at military camp. Yeah, there’s a built in reputation structure. It’s called rank. Students who are there longer generally have higher rank than new students. If you don’t, you’ve got a bad rep, and people know not to trust you. Trust is important in a military unit because it allows members to know who’s got their back, and who is useless. Squad leaders need to build trust in their squads and trust each member so that that team can be more effective. If you constantly need to baby-sit your soldiers, you don’t have an efficient team and you can’t adapt quickly (Bos, 135).
Desirable behavior amounted mostly avoiding gigs, things that docked your unit points. Also, it entailed helping others, even when helping others would make you vulnerable (Bos, 135). For example, I received a promotion to operation sergeant when I took time to ensure that the barracks was tidy prior to inspection, even though it meant that I received gigs myself. Trust, at least a military camp, was about putting the whole before yourself. This was the most desired behavior, and what the counselors deemed as reasons for promotions.
Clearly, promotions are the coveted award for good behavior. While carrying more responsibility and prestige, promotions also came with perks. First-class officers were able to carry sabers in parades, and second-class officers would carry sword. Commanders were given special mess privileges, and could order meals off campus twice per week. Officers could order meals off campus once per week. Those without positions were unable to do so. In addition, the higher rank meant better reputation, and newer soldiers trusted the higher ranked officers more. I remember that in my first year, one of the officers felt that I had been ranked too low among my class, and after discussing it with his superiors, he was able to increase my own rank, and thus my own reputation.
This is an interesting dichotomy. It means that those with a higher reputation can increase other’s ranks, which means that packs of friends may rise to the top, even if they are not the best candidates. This could be a problem with trust from other people, and may lower the overall efficiency of the unit. In addition, the officer’s food perks were often skirted by having officers order food for lower members for money. This meant that reputation, and more importantly trust, was abused.
Desirable behavior amounted mostly avoiding gigs, things that docked your unit points. Also, it entailed helping others, even when helping others would make you vulnerable (Bos, 135). For example, I received a promotion to operation sergeant when I took time to ensure that the barracks was tidy prior to inspection, even though it meant that I received gigs myself. Trust, at least a military camp, was about putting the whole before yourself. This was the most desired behavior, and what the counselors deemed as reasons for promotions.
Clearly, promotions are the coveted award for good behavior. While carrying more responsibility and prestige, promotions also came with perks. First-class officers were able to carry sabers in parades, and second-class officers would carry sword. Commanders were given special mess privileges, and could order meals off campus twice per week. Officers could order meals off campus once per week. Those without positions were unable to do so. In addition, the higher rank meant better reputation, and newer soldiers trusted the higher ranked officers more. I remember that in my first year, one of the officers felt that I had been ranked too low among my class, and after discussing it with his superiors, he was able to increase my own rank, and thus my own reputation.
This is an interesting dichotomy. It means that those with a higher reputation can increase other’s ranks, which means that packs of friends may rise to the top, even if they are not the best candidates. This could be a problem with trust from other people, and may lower the overall efficiency of the unit. In addition, the officer’s food perks were often skirted by having officers order food for lower members for money. This meant that reputation, and more importantly trust, was abused.
Assignment Seven - Reputation Systems (Katie Dreier)
People pride themselves on a positive reputation. In a sorority, a good reputation tends to beget many friends who will respect and trust you. According to the Bos article, “without trust, partners will not share information openly” (135). Those who trust each other are more likely to form close friendships and a tight knit group as they tend to act cooperatively in a group environment and thereby “ensure maximum group benefit” (Bos 136). Therefore, trust can be defined as “a willingness to be vulnerable, based on positive expectations about the actions of others.” These positive expectations are confirmed by a person’s reputation based on their previous actions. A system can help confirm a person’s reputation by keeping track of people’s perceptions of the person’s past behaviors. According to Resnick, “a reputation system collects, distributes, and aggregates feedback about participants’ past behavior.” Simply put, when past behavior gets added up and people notice a positive trend, a good reputation forms and trust increases. A popular characteristic of sorority life is the chance to form the strong bonds of a sisterhood among your sisters. Trust is a key ingredient in guaranteeing this bond.
In my sorority’s pledge class, we have a reputation system that is not based solely on our personalities, rather on our actions. We use a point system. Sisters are rewarded for desirable behaviors such as attending events, dedicating hours to philanthropy events, and doing favors for sorority sisters through points. Sisters who no longer need to earn points monitor the system. This ensures that someone greedy for more points will not manipulate the system. The ranking of points is a well-balanced, well-respected system among the sisters. A large amount of points is directly related to how involved you are in the house. Because the system is not based on our personalities, sisters with less points are not disrespected, it merely means they may not have as tight bonds with other sisters as they probably haven’t spent as much time in the house and at events with other sisters. This FtF time that they’ve missed out on “decreases social distance … and makes trust agreements easier to form and maintain” and will make it harder to build trust with the other sisters (Bos 135).
In my sorority’s pledge class, we have a reputation system that is not based solely on our personalities, rather on our actions. We use a point system. Sisters are rewarded for desirable behaviors such as attending events, dedicating hours to philanthropy events, and doing favors for sorority sisters through points. Sisters who no longer need to earn points monitor the system. This ensures that someone greedy for more points will not manipulate the system. The ranking of points is a well-balanced, well-respected system among the sisters. A large amount of points is directly related to how involved you are in the house. Because the system is not based on our personalities, sisters with less points are not disrespected, it merely means they may not have as tight bonds with other sisters as they probably haven’t spent as much time in the house and at events with other sisters. This FtF time that they’ve missed out on “decreases social distance … and makes trust agreements easier to form and maintain” and will make it harder to build trust with the other sisters (Bos 135).
Assignment 7: Christina Caiozzo
In the summers, I teach swimming lessons to little kids. It is important to develop a reputation system for this kind of community. We need to be able to differentiate the swimmers by skill level so we know who the weak and strong swimmers are. Giving the swimmers tests and then dividing them up by groups according to skill is a good way to do this. They might have to: tread water, swim different strokes, swim to one end of the pool and back, or even swim 20, 50, or 100 laps. Those that have the most basic skills would be put in a beginner group, and wear red bathing suits. Intermediate levels could wear green bathing suits, and the advanced could wear blue bathing suits.
Although Olsen et. al. (2002) are mainly interested in finding out whether face to face interaction is essential for developing trust, they do specify that “if higher degrees of trust can be established, organizations can work more efficiently, and adapt more quickly to changing circumstances” (p. 135). This principle applies to the swimming community. If the instructors can trust that the swimmer’s group matches his/her ability level, than their attention can be distributed accordingly. Likewise, the swimmers need to be able to trust the instructor’s advice and guidance in order to progress.
Desirable behaviors are those that reinforce the reputation system. If swimmers look forward to moving up the levels and receiving a new bathing suit, then they will work hard in their lessons in order to pass the test to get to the next level. If swimmers base their impressions of other swimmers’ skills on the color of their bathing suit, that would also be desirable. This way, swimmers can be aware of who should be in the deep end, etc.
Resnick et. al. (2000) notes that it is important to distribute feedback to the swimmers about their performance in order to “inspire an expectation of future interaction”(p.47). Rewarding the swimmers by giving them a patch or ribbon on their bathing suit every time they do something noteworthy would hopefully encourage them to keep up their performance. It is important to remember not to rate the swimmer themselves, but the activity they accomplish.
Manipulation of the system would be possible. Since instructor feedback would be subjective, some instructors may provide ribbons for behavior that other instructors would not deem commendable. Also, it is entirely possible for swimmers not affiliated with swimming class to wear a red, blue or green bathing suit. This would be confusing to the lifeguards, the other swimmers, and the reputation system. Additionally, since swimming lessons are only during the summer, it would be possible for a swimmer to come in late, and realize that their behavior will not matter since the summer will end before there’s a chance to move up a new level. This may decrease motivation, and make the reputation system ineffective.
Although Olsen et. al. (2002) are mainly interested in finding out whether face to face interaction is essential for developing trust, they do specify that “if higher degrees of trust can be established, organizations can work more efficiently, and adapt more quickly to changing circumstances” (p. 135). This principle applies to the swimming community. If the instructors can trust that the swimmer’s group matches his/her ability level, than their attention can be distributed accordingly. Likewise, the swimmers need to be able to trust the instructor’s advice and guidance in order to progress.
Desirable behaviors are those that reinforce the reputation system. If swimmers look forward to moving up the levels and receiving a new bathing suit, then they will work hard in their lessons in order to pass the test to get to the next level. If swimmers base their impressions of other swimmers’ skills on the color of their bathing suit, that would also be desirable. This way, swimmers can be aware of who should be in the deep end, etc.
Resnick et. al. (2000) notes that it is important to distribute feedback to the swimmers about their performance in order to “inspire an expectation of future interaction”(p.47). Rewarding the swimmers by giving them a patch or ribbon on their bathing suit every time they do something noteworthy would hopefully encourage them to keep up their performance. It is important to remember not to rate the swimmer themselves, but the activity they accomplish.
Manipulation of the system would be possible. Since instructor feedback would be subjective, some instructors may provide ribbons for behavior that other instructors would not deem commendable. Also, it is entirely possible for swimmers not affiliated with swimming class to wear a red, blue or green bathing suit. This would be confusing to the lifeguards, the other swimmers, and the reputation system. Additionally, since swimming lessons are only during the summer, it would be possible for a swimmer to come in late, and realize that their behavior will not matter since the summer will end before there’s a chance to move up a new level. This may decrease motivation, and make the reputation system ineffective.
Assignment 7: Melanie Aliperti
A few months ago, one of my friends that was participating in a prank war with some other friends woke up to people ringing her doorbell telling her they were there to pick up the free couch. Apparently her friends had posted an advertisement for a “free furniture takeaway” on Craig’s List. This leads me to believe that as useful as Craig’s List is, it could use a little help in the reputation system department. In a website that focuses a lot around organizing purchases, exchanges, advertising events, requesting services and tons of other person-to-person transactions, trust is clearly necessary. Forums like Craig’s List are subject to a lot of fraud, and a better reputation system would help people feel more comfortable about using the site.
I think a type of “Feedback Forum” like Resnick et al. would be very useful for Craig’s List. Users should be allowed to rate one another based on how good of a seller, buyer, searcher, etc. they are. In addition each of these areas should have subcategories that suggest how true to description items they were selling were, how fast they were with exchanges, how quickly they paid, etc. In addition, I think a feature that allows video or audio chat between users would be useful since the Bos et al. study suggests that trust fragility is less of an issue with richer mediums. In addition user’s could be required to upload profiles with pictures, since reduced social presence can often foster trust fragility problems.
One possible reward tactic would be to offer users with good ratings higher listing results on searches. If I search for Mets tickets, the users offering them with high ratings’ posts could come up first, therefore giving people an incentive to get good feedback. The system isn’t perfect however and I think would be especially susceptible to the pseudonym issues that Resnick et al. discuss. Perhaps in order to create a Craig’s List account, you should be required to give your full name, phone number, address and email information, then when people create new accounts, they should check their information to see if it’s valid and compare it to other’s in the database to make sure people aren’t creating new account to avoid bad ratings.
It’s not a perfect solution, but it might have possibly prevented my friend from waking up to someone attempting to haul away her whole living room.
I think a type of “Feedback Forum” like Resnick et al. would be very useful for Craig’s List. Users should be allowed to rate one another based on how good of a seller, buyer, searcher, etc. they are. In addition each of these areas should have subcategories that suggest how true to description items they were selling were, how fast they were with exchanges, how quickly they paid, etc. In addition, I think a feature that allows video or audio chat between users would be useful since the Bos et al. study suggests that trust fragility is less of an issue with richer mediums. In addition user’s could be required to upload profiles with pictures, since reduced social presence can often foster trust fragility problems.
One possible reward tactic would be to offer users with good ratings higher listing results on searches. If I search for Mets tickets, the users offering them with high ratings’ posts could come up first, therefore giving people an incentive to get good feedback. The system isn’t perfect however and I think would be especially susceptible to the pseudonym issues that Resnick et al. discuss. Perhaps in order to create a Craig’s List account, you should be required to give your full name, phone number, address and email information, then when people create new accounts, they should check their information to see if it’s valid and compare it to other’s in the database to make sure people aren’t creating new account to avoid bad ratings.
It’s not a perfect solution, but it might have possibly prevented my friend from waking up to someone attempting to haul away her whole living room.
Brianne Wingate Reputation System
I am part of a technological consulting project for my sorority's international office. Representatives (comprised mostly of alumnae plus two collegians including myself) are spread throughout the country, so meetings primarily take place through Skype. Trust matters in this community because each member has to satisfactorily fulfill her role in order for us to make any progress on our undertakings- if one of my colleagues says that she will e-mail me information by a certain time, I need to be able to make solid plans assuming that she will keep her promise. For example, we are currently working on a system that would give chapters a more user-friendly digital master calendar. However, we have to make sure that everyone submits their input before moving forward. If we have to keep waiting for one member to comment, we will never complete the task.
Desirable behaviors include active participation in synchronous meetings as well as online threads. We want each member to apply her knowledge of web-based interfaces in a way that would make Delta Gamma's online efforts as easy and proficient as possible to the greatest number of users. My reputation system would reward active participation with anchor-shaped icons (since an anchor is my sorority's symbol) next to the users' Skype names and e-mail addresses when they are communicating with other group members. Skype would monitor the user's voice during conference calls, and e-mail would keep track of the user reading the thread and the number of responses sent. The respective clients would tally up the number of utterances or responses compared to the total amount of communication among the group. There would be four levels of participation- white, bronze, blue, and pink. The least amount of participation (zero) would be represented by white, whereas the highest would be represented by pink.
This scheme could be easily manipulated- users could make nonsense utterances during Skype meetings or send one-word responses over e-mail. They could also refuse to take part in the system all together so that they can participate as much or as little as they would like. However, these manipulations would be clear to the group and would encourage the group to keep their own notes about participation in addition to the automated counts.
Desirable behaviors include active participation in synchronous meetings as well as online threads. We want each member to apply her knowledge of web-based interfaces in a way that would make Delta Gamma's online efforts as easy and proficient as possible to the greatest number of users. My reputation system would reward active participation with anchor-shaped icons (since an anchor is my sorority's symbol) next to the users' Skype names and e-mail addresses when they are communicating with other group members. Skype would monitor the user's voice during conference calls, and e-mail would keep track of the user reading the thread and the number of responses sent. The respective clients would tally up the number of utterances or responses compared to the total amount of communication among the group. There would be four levels of participation- white, bronze, blue, and pink. The least amount of participation (zero) would be represented by white, whereas the highest would be represented by pink.
This scheme could be easily manipulated- users could make nonsense utterances during Skype meetings or send one-word responses over e-mail. They could also refuse to take part in the system all together so that they can participate as much or as little as they would like. However, these manipulations would be clear to the group and would encourage the group to keep their own notes about participation in addition to the automated counts.
Subscribe to:
Posts (Atom)
Posts
