Leveraging editor collaboration patterns in wikipedia

Abstract

Predicting the positive or negative attitude of individuals towards each other in a social environment has long been of interest, with applications in many domains. We investigate this problem in the context of the collaborative editing of articles in Wikipedia, showing that there is enough information in the edit history of the articles that can be utilized for predicting the attitude of co-editors. We train a model using a distant supervision approach, by labeling interactions between editors as positive or negative depending on how these editors vote for each other in Wikipedia admin elections. We use the model to predict the attitude among other editors, who have neither run nor voted in an election. We validate our model by assessing its accuracy in the tasks of predicting the results of the actual elections, and identifying controversial articles. Our analysis reveals that the interactions in co-editing articles can accurately predict votes, although there are differences between positive and negative votes. For instance, the accuracy when predicting negative votes substantially increases by considering longer traces of the edit history. As for predicting controversial articles, we show that exploiting positive and negative interactions during the production of an article provides substantial improvements on previous attempts at detecting controversial articles in Wikipedia.

Identifying Controversial Wikipedia Articles Using Editor Collaboration Networks

“…From the previous work, we considered the following controversy detection methods: -Mutual Reverts [Sumi et al 2011]: this method uses a simple score based on the number of (pairs of) edits of that are mutually reverted by the same editors. The inherent assumption here is that reverts that are not replied back by the other editor are due to fixing vandalism instead of disagreement.…”

“…To build the ground truth, we randomly picked 240 controversial articles from the list of articles manually identified as such by the Wikipedia community. 8 We disregard dispute tags in our choice, as they do not seem a reliable predictor of controversy; at the very least, the absence of such tags does not mean lack of controversy [Sumi et al 2011]. In fact, out of the 240 controversial articles in our sample, only 122 received a dispute tag at some point in their history.…”

“…Permissions may be requested from Publications Dept., ACM, Inc., 2 Penn Plaza, Suite 701, New York, NY 10121-0701 USA, fax +1 (212) 869-0481, or permissions@acm.org. c 2015 ACM 2157-6904/2015/03-ART5 $15.00 DOI: http://dx.doi.org/10.1145/2630075 complementary perspectives, strong differences among editors can cause an article to become untrustworthy or lead to so-called edit wars [Sumi et al 2011]. …”

“…However, previous attempts to solve the problem have relied mostly on simple heuristics that either completely ignore editor interactions or rely on a limited set of basic edit operations such as reverts or deleted words [Brandes et al 2009b;Vuong et al 2008;Kittur et al 2007;Sumi et al 2011]. In practice, however, these simplistic approaches are not robust in detecting controversy in Wikipedia .…”

Quantifying trust dynamics in signed graphs, the S-Cores approach

“…Indeed, several social applications support and publish such links (e.g., Epinions and Slashdot). Furthermore, recent research tackled the problem of extracting implicit signed networks, through the analysis of user interactions in the application [14,13,3,18]. This paper deals with the issue of reciprocity in signed graphs and introduces reciprocity in signed networks both as a local and a global property illustrating the value of the latter.…”