Research And Application Of Question Answering System In Automobile Field Based On Knowledge Graph

With the development of science and technology,the amount of information generated in modern society is increasing exponentially.How to quickly and effectively mine effective information from massive information has become a key research topic at present.The emergence of knowledge graph technology has largely solved this problem.Knowledge graph can structure the information between data and reveal the relationship between information.Knowledge Graph has been favored by many industries by virtue of its powerful information construction capabilities.With the development of natural language processing technology,knowledge graphs have been applied in many fields,including intelligent question answering,recommendation systems,etc.With the improvement of living standards,people’s demand for online car information consultation has increased exponentially,and the workload of manual customer service has greatly increased.In order to solve this problem,we constructs a question answering system based on the knowledge graph of the automotive domain.The main tasks are as follows:First,we use Python language,combined with Beautifulsoup+Requests+Selenium technology to crawl the relevant data information of the automotive industry portal website,and build a knowledge graph in the automotive field after scripting and manual cleaning;Then,according to the characteristics of Chinese expression and the naming rules of entities in the automotive field,the BBA-CRF(BERT(wwm)+Bi LSTM+Attention+CRF)algorithm model in this system is proposed to realize entity recognition in the knowledge graph;In addition,the B-CNN(BERT(wwm)+CNN)algorithm model is proposed to realize the identification and extraction of massive relationships in the knowledge map of the automotive domain;Finally,a question-and-answer system in the automotive field was built based on the MVC three-tier architecture,and simulation tests were carried out,which proved that the system in this paper can accurately answer user and automotive-related questions.