CRQA: Credibility Retrieval for Medical Question Answer Service

With the rapid development of artificial intelligence and its wide application on the ground, more and more devices is embedded in intelligent question answering technology. The development of mobile Internet technology and the popularization of community platform provide us with huge and jumbled data. Credibility has become an urgent problem to be solved in question answering system. As far as the medical platform is concerned, there is a huge noise problem in the current general medical information retrieval. Based on this, we propose a trusted evaluation mechanism based on the information of the medical community platform. In order to search and select more accurate and credible answers, we propose an optimization method of search ranking based on credible evaluation mechanism. By analyzing the professional competence of information providers, this method trustfully ranks multiple answers in the question-and-answer set of medical communities, which largely solves the interference of “one-question multi-answer” phenomenon in the question-and-answer system on retrieval. At the same time, it uses department classification, subdivides medical directions, and further matches the areas of concern of information providers, so as to improve the accuracy and hit rate of the retrieval system.