Research And Implementation Of Film And Television Knowledge Question Answering System Based On Knowledge Graph

With the development of the national economy,people’s demand for spiritual culture is increasing,which promotes the rapid development of Chinese film and television industry.At present,there are a large number of works in Chinese film and television industry,with different styles and splendid scenes.However,the rapid growth in the number of film and television works has brought about a series of contradictions,especially the contradiction between the accurate demand of users’ film and television information and the fuzzy search results of web pages.Users’ search for film and television related information mainly depends on the search of film and television web pages,which leads to the inaccurate results returned to users and the unfriendly search process.Therefore,it is particularly important to realize a scientific and effective intelligent question and answering system to provide users with the accurate inquiry of film and television knowledge to adapt to the rapid development of the film and television industry.A knowledge graph of the film and television field with the accurate information and new content is constructed.A knowledge question-and-answer model of the film and television field is proposed,and a film and television knowledge question-and-answer system based on knowledge graph is designed and implemented in this dissertation.The specific research work is as follows:(1)The knowledge graph of film and television is constructed.Based on the existing domain ontology construction technology,a construction method suitable for movie domain ontology is proposed,and a movie domain ontology oriented to question-and-answer scenes is constructed in this dissertation.By designing and implementing a web crawler system to acquire knowledge in the field of film and television,and using the mainstream graph database to store knowledge in the field of film and television,a reasonable structure and application-oriented knowledge graph in the field of film are constructed.(2)A knowledge question and answer model in the field of film and television is proposed.The model is composed of a question keyword extraction module,question classification module,question query module and human-computer interaction module.In the question keyword extraction module,we use the film question keyword extraction model via ALBRET-Bi LSTM-CRF to get a good model effect after training.In the question classification module,through the construction of the film question and answer domain data set,the film question classification template is designed,and the film question classification model via ALBRET-Text CNN is adopted,which achieves a good classification effect.In the problem query module,Cypher query is designed according to the classification template of movie and television problems,which makes the graph database Neo4 j used to efficiently complete complex query tasks.From the user’s point of view,the Web dialog window is determined as the interaction mode in the human-computer interaction module.(3)The film and television knowledge question and answer system are designed and implemented.Based on the classical hierarchical architecture model of domain-driven design,a hierarchical architecture model adapted to the question answering system is proposed,and the overall architecture of the film and television question answering system is designed.According to the design,the film and television knowledge question and answer system based on knowledge graph are realized.After the test,the system can effectively complete the user’s film and television knowledge of the question task.