Skip to Main Content
Community-driven Question Answering (CQA) services on the Internet enable users to share content in the form of questions and answers. Nonetheless, questions usually attract multiple answers of varying quality from the user community. For this reason, the purpose of this paper is to identify high quality answers from a group of candidate answers obtained from semantically similar questions that match with the new question. To do so, a three-component quality framework comprising social, textual and content-appraisal features of user-generated answers in CQA services was developed and tested on Yahoo! Answers. The results of our logistic regression analysis revealed content-appraisal features to be the strongest predictor of quality in user-generated answers. Specifically, these features include dimensions such as comprehensiveness, truthfulness and practicality. Going forward, the quality framework developed in this paper may be used to pave the way for more robust CQA services.