Research on Enhancing Generalization Ability of Question Answering System by Retelling Technology

Retelling generation technology has important applications in many aspects, such as: in question answering system, it can be used to improve the system’s ability to understand questions and optimize the system performance; With the continuous development and maturity of various underlying technologies of natural language processing, it is possible to study retelling, which has attracted more and more attention. In this paper, a method of generating retelling questions based on template matching is proposed, which can effectively preserve the structural and semantic features of questions. Introducing functional tags to highlight the structural features of questions; The introduction of dependency relationship improves the generalization performance of question template; By introducing candidate sorting, the accuracy of the generated results is greatly improved. As the retelling technique mainly operates on monolingual language, it is possible to use large-scale monolingual language data to generate retelling, thus alleviating the problem of insufficient bilingual language data to a certain extent.