标题：Research on the quality prediction of online Chinese question answering community answers based on comments
作者：He, Xun ;Wang, Lianhai ;Zhang, Weinan ;Zhang, Peijun
作者机构：[He, Xun ;Wang, Lianhai ;Zhang, Weinan ;Zhang, Peijun ] Qilu University of Technology(Shandong Academy of Science), Shandong Computer Science Center(N 更多
会议名称：2nd International Conference on Big Data Technologies, ICBDT 2019
会议日期：August 28, 2019 - August 30, 2019
来源：ACM International Conference Proceeding Series
摘要： © 2019 Association for Computing Machinery.">With the rapid development of online Community Question Answer (CQA), a large volume of valuable data has been accumulated in CQA sites, as well as a huge number of low-quality answers. To improve the user-friendliness of CQA sites and help users find high-quality answers quickly, in this paper, we propose a supervised learning model to evaluate the quality of answers on Chinese CQA sites. We build a quality evaluation model based on the pairwise learning-to-rank algorithm and combine a set of features to rank answers according to their quality. In the quality evaluation process, we also propose an innovative type of feature named "the sentiment polarity of comments". Experimental results on real CQA data show that our method can efficiently produce a quality-ranking list of answers. Moreover, the proposed sentiment polarity feature can improve the performance of the quality evaluation model significantly.
© 2019 Association for Computing Machinery.