Research And Application Of Knowledge Question Answering Model For Confucius Family Tree

China is a unified and multi-ethnic country.Its long history has accumulated a lot of brilliant and excellent cultures,and genealogy is an irreplaceable part of them.Genealogy not only records the rise and fall of a family,but also reflects the historical and cultural background of the society at that time.It is a kind of humanistic data with unique functions.The Genealogy of Confucius records the information of the lineage and reproduction of the people of Confucius as their ancestor,and carries the spirit and soul of Chinese Confucian culture.Therefore,it is of great research value and practical significance to explore the semantic representation,semantic reasoning and application of The Genealogy of Confucius.Knowledge base question answering system is an intelligent platform that integrates structured and unstructured knowledge extraction,representation,storage and retrieval,realizes knowledge-based semantic analysis and cognitive reasoning,and solves users' cognition and decision-making problems in a simple dialogue mode.Aiming at the question answering system of the genealogy of Confucius,this paper applies the technology and architecture of knowledge graph,deep learning and graph database,and conducts the following research work in three aspects: knowledge graph construction,knowledge base question answering model design and question answering system development:(1)This paper puts forward the ontology construction method of genealogy,and constructs the knowledge graph of the genealogy of Confucius.Firstly,based on the analysis of the original data of The Genealogy of Confucius,an improved ontology construction method is proposed and the initial ontology structure of the genealogy is constructed.Then,according to ontology constraint conditions,specific information extraction methods are developed for text data with different structures.Finally,the extracted effective information is integrated and stored in the Neo4 j graph database,which lays a knowledge foundation for the construction and application of knowledge base question answering model and system.(2)A knowledge base question answering model combining deep learning and template was proposed,and a knowledge base question answering strategy based on semantic parsing of genealogy was designed.Firstly,in order to reduce the user's waiting time,a sub-module of question judgment is added into the semantic parsing module.Secondly,in order to improve the accuracy of semantic parsing,Albert-Text CNN model and Albert-BLSTM-CRF model,which are introduced into the pre-trained language model,are proposed for question judgment and entity extraction respectively.Experimental results show that,compared with the traditional model,the accuracy of question judgment and entity recognition is increased by1.34% and 2.01%,respectively,and the accuracy of knowledge base question answering model is 92%.(3)The question answering system of the genealogy of Confucius is designed and developed.In order to conduct research on the knowledge graph of the genealogy of Confucius effectively and popularize the culture of Confucius genealogy and Confucian culture extensively,a question answering system was implemented based on the knowledge graph of the genealogy of Confucius and the knowledge base question answering model.In accordance with the software engineering development process,the system uses Flask as the development framework,uses cytoscape.js library for Web visualization of the graph,and realizes the function of visual retrieval and knowledge question-and-answer for the contents of the first set of the genealogy of Confucius.In conclusion,the genealogy of Confucius based on the knowledge graph provides new methods and ideas for the study of Confucian culture,and the question answering system based on the knowledge graph explores new forms and ways for the inheritance and promotion of Chinese Confucian culture.