Finding Experts in Community Question Answering Based on Topic-Sensitive Link Analysis

Abstract

Community question answering(CQA) websites such as Quora and StackOverflow provide a new way of asking and answering questions which are not well served by general web search engines. With the huge volume and ever-increasing number of users and questions, effective strategies of ranking experts for different questions need to be proposed. In this paper, we first make some analysis on the network structure of the CQA website. Based on these works, we further propose an expert finding method NEWHITS, which considers the topical similarity of the users and can well adapts to the feature of the CQA. Then, we apply the NEWHITS algorithm to user authority ranking. The comparison experiments with StackOverflow data are conducted and the experimental results demonstrate that the method we proposed performs better than traditional link analysis methods in the user authority ranking.