A Corpus of Wikipedia Discussions: Over the Years, with Topic, Power and Gender Labels

A Corpus of Wikipedia Discussions: Over the Years, with Topic, Power and Gender Labels Vinodkumar Prabhakaran∗, Owen Rambow‡ ∗ Department of Computer Science, Stanford University, California, CA 94305, USA. ‡ Center for Computational Learning Systems, Columbia University, New York, NY 10027, USA. vinod@cs.stanford.edu, rambow@ccls.columbia.edu Abstract In order to gain a deep understanding of how social context manifests in interactions, we need data that represents interactions from a large community of people over a long period of time, capturing different aspects of social context. In this paper, we present a large corpus of Wikipedia Talk page discussions that are collected from a broad range of topics, containing discussions that happened over a period of 15 years. The dataset contains 166,322 discussion threads, across 1236 articles/topics that span 15 different topic categories or domains. The dataset also captures whether the post is made by an registered user or not, and whether he/she was an administrator at the time of making the post. It also captures the Wikipedia age of editors in terms of number of months spent as an editor, as well as their gender. This corpus will be a valuable resource to investigate a variety of computational sociolinguistics research questions regarding online social interactions. Keywords: computational sociolinguistics, Wikipedia, online interactions, power, gender, dialog 1. Introduction Computational analysis of online social interactions has become an active field of research in recent years. Re- searchers have studied the linguistic and dialogic patterns of these interactions, the network structures they form, as well as how these patterns and structures relate to the social relations that exist between the interactants. These studies cover a wide range of genres such as social networking websites, email interactions, and online discussion fo- rums. Within the genre of online discussion forums, the discussions happening in Wikipedia Talk pages (forums where Wikipedia editors discuss and debate the edits to the Wikipedia articles) have garnered special attention due to the fact that Wikipedia Talk is one of the very few online sources for task-oriented interactions. In this paper, we present a large corpus of Wikipedia Talk page discussions that are collected from a broad range of topics, containing discussions that happened over a pe- riod of 15 years. The dataset contains 166,322 discus- sion threads, across 1236 articles/topics that span 15 differ- ent topic categories or domains. The dataset also captures whether the post is made by an registered user or not, and whether he/she was an administrator at the time of mak- ing the post. It also capture the Wikipedia age of editors in terms of number of months spent as an editor, as well as their gender. This corpus will be a valuable resource to investigate a variety of computational sociolinguistics re- search questions regarding online social interactions. 2. Related Work There is a wide array of computational studies analyzing the dynamics of the collaborative editing process of building Wikipedia. One line of work focuses mainly on meta information such as history of edits, deletes, reverts, and dispute tags …

Content truncated to 500 words