Finding others online: reputation systems for social online spaces

From AcaWiki
Jump to: navigation, search

Citation: Carlos Jensen, John Davis, Shelly Farnham (2002) Finding others online: reputation systems for social online spaces. Proceedings of the SIGCHI conference on Human factors in computing systems: Changing our world, changing ourselves (RSS)
DOI (original publisher): 10.1145/503376.503456
Semantic Scholar (metadata): 10.1145/503376.503456
Sci-Hub (fulltext): 10.1145/503376.503456
Internet Archive Scholar (fulltext): Finding others online: reputation systems for social online spaces
Tagged: Computer Science (RSS) CSCW (RSS), Social Computing (RSS)

Summary (Abstract)

Jenson, Davis, and Farnham (2002) is a letter published in CHI 2002 that presents the results of a survey based exploration done largely from with in the Microsoft Social Computing group on what might drive useful reputation systems for the design of online communities.

The paper frames its exploration in terms of the need for reputation systems in order to help users engage in more meaningful and helpful interactions using social computing systems and, through this process, to raise the quality of discussion from the perspectives of most users (i.e., reduce flaming and bad behavior or render it less disruptive).

Reputation seems very clearly limited to describing the quality of other users of an online system and not to content, submissions, or products. The paper presents an ontology of reputation based systems that includes 5 parts:

  1. Ranking systems (essentially automated reputation systems)
  2. Rating systems (where humans provide the rankings and feedback of each other)
  3. Collaborative filtering systems (which weight ratings based on agreement between the user and the rater)
  4. 'Implicit peer-based reputation systems (essentially rankings based largely our friends' behaviors)
  5. Explicit peer-based reputation systems (ratings, ideally from our friends, with social network information provide alongside)

The authors offer a hypotheses that for more social based tasks (e.g., selecting a person to chat with), these social and peer-based information systems will be preferred by users.

The authors test this with a set of two surveys and with a questionnaire. They find that when selecting a chat partner, similarity of interest is most important and that friends opinions come immediately after. They find similar results (i.e., a focus on similarity information and ratings by friends) as important in their subsequent work.

In their conclusion, they present a user test which shows diagrams of social networks which users seem to be able to make sense out of and suggest that these types of interfaces may be useful for social networking systems.