Research on Methods and Applications Related to Question-and-Answer Dialogue Systems

Abstract

In the face of more and more network data information, search engines have gradually become the main retrieval method to obtain relevant information knowledge. However, in today's increasingly explosive development of information on the Internet, by contrast, traditional search engines have problems such as semantic understanding and complicated answers. Therefore, question answering systems are more important. The automatic question answering system generally adopts natural language processing related technologies. When users ask questions, the system automatically judges and gives answers. It involves computer linguistics, machine learning, artificial intelligence and other popular technology research. According to different classification criteria, the automatic question answering system is roughly divided into open field automatic question answering system and stereotyped automatic question answering system.. This thesis investigates methods and applications related to question-and-answer dialogue systems. On the methodological side, we introduce commonly used datasets and the principles and techniques of text, speech and visual question and answer systems, and analyse in detail the excellent example ChatGPT. In terms of applications, we present the application of Q&A dialogue systems in search engines, smart campuses. There is some reference value.