A Lightly Supervised Approach to Role Identification in Wikipedia Talk Page Discussions

Authors

  • Oliver Ferschke Carnegie Mellon University
  • Diyi Yang Carnegie Mellon University
  • Carolyn Rosé Carnegie Mellon University

DOI:

https://doi.org/10.1609/icwsm.v9i5.14691

Keywords:

wikipedia, discourse analysis, talk page discussions, nlp, role identification, machine learning

Abstract

In this paper we describe an application of a lightly supervised Role Identification Model (RIM) to the analysis of coordination in Wikipedia talk page discussions. Our goal is to understand the substance of important coordination roles that predict quality of the Wikipedia pages where the discussions take place. Using the model as a lens, we present an analysis of four important coordination roles identified using the model, including Workers, Critiquers, Encouragers, and Managers.

Downloads

Published

2021-08-03

How to Cite

Ferschke, O., Yang, D., & Rosé, C. (2021). A Lightly Supervised Approach to Role Identification in Wikipedia Talk Page Discussions. Proceedings of the International AAAI Conference on Web and Social Media, 9(5), 43-47. https://doi.org/10.1609/icwsm.v9i5.14691

Issue

Vol. 9 No. 5 (2015): ICWSM Workshop Technical Report WS-15-19 (Wikipedia, a Social Pedia: Research Challenges and Opportunities)

Section

Wikipedia, a Social Pedia