The Research of Multi-label Question Classification in Community Question Answering

The internet in the web2.0 mode. In Yahoo! Answers, Baidu Knows and Sina to ask, as the representative of the Community Question Answering. Through the remarkable features open and flexible user interactive features, attracting a large number of users, and gradually become popular user's favorite knowledge sharing platform. Compared with traditional search engines, Community Question Answering can better express the needs of users and provide users with personalized service. Question classification is an important part of the Community Question Answering. Classification accuracy directly affects the quality of personalized service. Aiming at these questions in CQA, we propose an algorithm for Multi-label question classification. The algorithm fully takes into account the correlation among labels on the basis of characteristics extension, iterating the algorithm of Multi-label question classification. Experiments show that this algorithm can effectively solve the multi-label question classification problem in CQA.