Predicting Best Answerers for New Questions: An Approach Leveraging Topic Modeling and Collaborative Voting

Abstract

Community Question Answering (CQA) sites are becoming increasingly important source of information where users can share knowledge on various topics. Although these platforms bring new opportunities for users to seek help or provide solutions, they also pose many challenges with the ever growing size of the community. The sheer number of questions posted everyday motivates the problem of routing questions to the appropriate users who can answer them. In this paper, we propose an approach to predict the best answerer for a new question on CQA site. Our approach considers both user interest and user expertise relevant to the topics of the given question. A user’s interests on various topics are learned by applying topic modeling to previous questions answered by the user, while the user’s expertise is learned by leveraging collaborative voting mechanism of CQA sites. We have applied our model on a dataset extracted from StackOverflow, one of the biggest CQA sites. The results show that our approach outperforms the TF-IDF based approach.