Question-answering system is a high and particular form of retrieval system. Itautomatically answers a question posed in natural language. The research of QAbecomes a hot spot mainly due to the ever-increasing demand of fast and accurateinformation gaining. The traditional Question-answering system has a lot of deficienciesin information organizing and user interaction, so how to design a question-answeringsystem with not only good Q/A functions but also some SNS characteristics becomesthe main job of this paper.In this paper, the research status of both domestic and foreign question-answeringsystem is firstly discussed. Then a SNS Question-answering system model is designedbased on the traditional question-answering system. It links users through theinformation which they are all somehow connected. Secondly, the related theories andtechnologies of each processing phase of a question-answering system are discussed,mainly focusing on SNS, inverted index, HowNet similarity calculation, etc. Thecalculations between questions are done through words calculations based on HowNet.Finally, a SNS question-answering system especially for remote education is presentedand implemented and the result shows that the system performs well and possessescertain SNS characteristics. |