Social Rewarding

Wikipedia - the most famous free encyclopedia - has been grown to the biggest wiki community site where hundreds of thousands of users all around the world post and edit articles in many different languages. As other online communities have troubles motivating users to participate actively, the tremendous contribution rate on Wikipedia has led to many problems like wrong information, copyright violations, or users' misbehavior, for example, spammers or trolls. In this project, we are going to present techniques where the fundamental problem of both - reaching a critical mass of active users - is addressed.

On the one hand, Wikipedia has the problem that published information is not checked for its accuracy and legality. There has to be a big and heavy involved community which is cross-checking and proofing information for its correctness. However, the operators of Wikipedia have not only a social but also a legal responsibility to publish only correct and faultless information to assure their creditability. On the other hand, many online communities have troubles motivating enough users to build an "active" community. As participation of members is the key factor for a successful online community, good motivating factors are essential. With this work we are going to focus on an approach to motivate users to participate in an online community by making use of developed social rewarding techniques.

In an online community, social rewarding is in the majority of cases based on accentuation of the most active members. As money cannot be used as a motivating factor, there have to be others like status, power, acceptance, or glory. We explain different social rewarding methods which aim to meet these needs of users.

The techniques presented are focussing primarily on automatic investigations of quantitative and qualitative characteristics of published articles. As a proof of concept, four social rewarding mechanisms were implemented using the software MediaWiki (which is also used by Wikipedia):

• Amount of References - This social rewarding method uses Google's search API to build an index quality number based on three different criteria: the size of a reference, the number of links pointing to this reference, and the number of links pointing to the specific article.

• Rating of Articles - A user centric evaluation of articles published on Wikipedia is still missing. We have implemented an open rating system where users can vote for or against an article (and optionally leave a comment) by making use of a predefined pointing scale.

• Most Viewed Articles - Visits of users are counted working with configured parameters. These first three social rewarding methods are applied to calculate a ranking of most productive authors.

• Recommender System - This technique does not touch the computation process for publishers. Here, it is possible to display recommended articles, revisions, authors, or authors with same interests. These statistics are calculated using data from other users in comparison with data collected for the profile of the logged-in user.

To generate a ranking of authors they are rated for their work and points are assigned using predefined criteria and automatically generated scales on the basis of all members of the whole community. These scores are then weighted according to the timeliness and the extent of a contribution. The weighting of all variables of the calculation process (~100) can be configured according to the behavior of the users and the needs of the online community.

For displaying results, various rankings of authors can be generated where the most active one will see her-/himself on the first place. To support shown results, two well-known data visualization techniques are used: Stars and Sparklines. A five star scaling (zero to five stars; six intervals) was implemented to display the overall contribution rate of a user. Sparklines are displayed beside a user's name to have an idea about the dispersal of the contributions. Here, sparklines are used to show the participation rate of a user over a certain period of time split by predefined intervals. We have chosen sparklines mainly because of their good integration in a context of words and their simplicity.

Our approach can be seen as a starting point to develop mechanisms to the important issue of motivating users to participate actively in a wiki system. Currently, we are planning to evaluate our implemented concepts in a larger setting.

Papers

Bernhard Hoisl: Motivate Online Community Contributions Using Social Rewarding Techniques - A Focus on Wiki Systems, Master Thesis, Vienna University of Technology, 2007.

Bernhard Hoisl, Wolfgang Aigner, Silvia Miksch: Social Rewarding in Wiki Systems – Motivating the Community. In: Online Communities and Social Computing, pp. 362–371. Springer, Berlin Heidelberg New York, 2007.

Bernhard Hoisl, Wolfgang Aigner, Silvia Miksch: Soziale Belohnung in Wiki Systemen. In: Wikis im Social Web – Wikiposium 2005/06, pp. 60–72. OCG – Austrian Computer Society, Vienna, 2007.

Bernhard Hoisl, Wolfgang Aigner: KonsumentInnen als ProduzentInnen - Partizipation als Erfolgsfaktor von Online Communities. In: TIMNEWS 03/2006, p. 5. Danube University Krems, 2007.