Implementation of Intelligent Question Answering System Based on Basketball Knowledge Graph
Ying Li, Jie Cao, Yongbin Wang · 2019
Currently most search engines query based on keywords or question-template matching. But for the retrieval about basketball or NBA, there are always too many feedback results, low accuracy and lack of intelligence. In this paper, an intelligent question answering system based on NBA basketball knowledge graph is implemented. Some methods are used in the question analysis module in the system, including question similarity calculation, named entity recognition, entity similarity calculation, and question-to-entity attribute mapping. In answer generation module of the system, a multi-strategy answer generation method is proposed. The experimental results show that our approach by combined with natural language processing technology and domain knowledge graph can well identify the input questions by user, and accurately feedback the answers to user queries.