Using Social Network to Predict the Behavior of Active Members of Online Communities

In this paper we use the social networks of active members (i.e. authors) in online communities to predict their future behavior. We limit our discussion to web forums, the most generic type of online communities. We first address the problem of measuring the ties between authors where there is (almost) no direct information (i.e. citations) about the tie between two authors. We developed a surrogate metrics to this problem so that a corresponding social network can be constructed. However, while such networks are suitable for describing authors' behavior, they are not very suitable for predicting their future behavior. Here we suggest a very specific method for creating networks of authors that is specially designed for prediction. The main characteristic of our approach is that we take into account all available data while putting more weight on more recent data (messages). In addition to networks of authors, we also use the two-mode network of threads (topics) to generate predictions.All network data and other information related to the posting habits of the authors are thus used to develop an algorithm that predicts the future behavior of the authors. The ultimate goal of our research is to generate explicit recommendations for authors of online communities about which authors and which topics/threads to follow. Finally, these predictions/recommendations can be used as the basis for a more user-friendly forum interface.

[1]  Lada A. Adamic,et al.  Friends and neighbors on the Web , 2003, Soc. Networks.

[2]  Robert F. Mills,et al.  Automatic Generation of Social Network Data from Electronic-Mail Communications , 2005 .

[3]  Marco Gonzalez,et al.  Author's Personal Copy Social Networks Tastes, Ties, and Time: a New Social Network Dataset Using Facebook.com , 2022 .

[4]  Bradley N. Miller,et al.  GroupLens: applying collaborative filtering to Usenet news , 1997, CACM.

[5]  Gediminas Adomavicius,et al.  Toward the next generation of recommender systems: a survey of the state-of-the-art and possible extensions , 2005, IEEE Transactions on Knowledge and Data Engineering.

[6]  Warren Sack,et al.  Conversation Map: An Interface for Very Large-Scale Conversations , 2000, J. Manag. Inf. Syst..

[7]  Greg Linden,et al.  Amazon . com Recommendations Item-to-Item Collaborative Filtering , 2001 .

[8]  Hsinchun Chen,et al.  Affect Analysis of Web Forums and Blogs Using Correlation Ensembles , 2008, IEEE Transactions on Knowledge and Data Engineering.

[9]  Susan C. Herring,et al.  Computer-mediated communication on the internet , 2005, Annu. Rev. Inf. Sci. Technol..

[10]  Lina Zhou,et al.  Social computing and weighting to identify member roles in online communities , 2005, The 2005 IEEE/WIC/ACM International Conference on Web Intelligence (WI'05).

[11]  Henry Tirri,et al.  Combining Topic Models and Social Networks for Chat Data Mining , 2004, IEEE/WIC/ACM International Conference on Web Intelligence (WI'04).

[12]  Mark H. Chignell,et al.  Identifying communities in blogs: roles for social network analysis and survey instruments , 2007, Int. J. Web Based Communities.

[13]  Kevin Crowston,et al.  Social Dynamics of FLOSS Team Communication Across Channels , 2008, OSS.

[14]  Kevin Crowston,et al.  Social dynamics of free and open source team communications , 2006, OSS.