Modeling Leading Users in Professional Social-network Based Community Question Answering Services ?

Differing from other traditional question-centric CQA services, professional social-network based CQA services lay a lot of emphasis on the quality of the content and social interactions within community. Besides questioning and answering, people pay more attention to authoritativeness and social reputation, especially some leading users. Others appreciate their contributions, and would like to make friends with them. Thus, in this study, we focus on identifying leading users within a specific topic area of professional user-centric CQA services. A leading user detecting model is proposed. Firstly, it extracts remarkable features of users’ leading ability, and then these features are exploited to train a classifier with multiple kernel learning (MKL) algorithms. Experimental results show the effectiveness of the proposed method.