Research on the method of knowledge base question answering

Knowledge base question answering is a way of communication between computer and human in the form of natural language, which is a branch of artificial intelligence. In recent years, knowledge base question answering has become a hot research and application direction in academia and industry. At present, question answering algorithms mainly include template-based approach, semantic analysis approach, deep learning approach. Many scholars have continuously improved existing algorithms for obtaining better accuracy. For giving people an overall understanding of the research significance, knowledge base and algorithm of question answering, in this study we mainly review three classical algorithms of knowledge base question answering, and make a comprehensive comparison of these three algorithms. We identified the deficiencies and challenges of knowledge base question answering. we propose future development directions for the principle of matching questions and answers to construct query sentences, deep learning models, and inherent features of language.