Research And Implementation Of Question And Answer Method For Earthquake Disaster Prevention And Based On Knowledge Graph

Earthquake disasters,as one of the most serious natural disasters threatening human society,pose a great threat to people’s lives and economic property.Frequent seismic activities and severe earthquake disaster prevention and control situation in China.The publicity and popularization of earthquake disaster prevention and control related knowledge is an important means to prevent earthquake disasters and reduce earthquake casualties and losses.Various sources of earthquake disaster prevention and control information on the Internet cannot guarantee the authority and accuracy of information,making it difficult to achieve precise knowledge queries,resulting in low efficiency of earthquake prevention and control information utilization,which is not conducive to the informationization development of earthquake prevention and control work.In order to effectively organize the use of many professional information in the field of earthquake prevention and control,and realize the high-quality service of earthquake prevention and control information,this paper uses knowledge modeling method to summarize the tasks of earthquake disaster prevention and control,comb the semantic relationships between knowledge,and then construct the knowledge graph of earthquake disaster prevention and control field,and focus on the key technology research of knowledge graph answering in earthquake disaster prevention and control field.A knowledge graph question answering method based on query path matching is proposed,and an earthquake disaster knowledge answering system is designed and built on this basis to achieve the structural organization and semantic service of earthquake disaster prevention and control information.The main research work of this paper is as follows:(1)Designed and built a knowledge graph of earthquake disaster prevention and control.By analyzing the characteristics of the disaster field and summarizing the key points of prevention and control work,we designed an ontology model of earthquake disaster prevention and control based on shelters,building anti-seismic and emergency shelter.On this basis,triple extraction was completed and visualized.(2)An improved question answering method for knowledge graph based on information retrieval is proposed.First,in order to improve the generalization of the question answering method,a pre-trained model is introduced to extract deep semantic information in the three key links of entity recognition,intent recognition,and semantic matching.Secondly,in order to effectively deal with complex problems,this paper designs ten types of query path generation rules.Finally,In order to avoid the problem that the traditional information retrieval question and answer method relies too much on entity recognition,a vector retrieval link is introduced as a system bottom-up method to improve the robustness of the system.Finally,after testing,the system reply accuracy rate reaches 87.9%.(3)Designed and implemented a prototype of knowledge-answering system in the field of earthquake prevention and control.Using the front-end and back-end separated development mode,I designed and implemented an intelligent question-answering system for earthquake disaster prevention and control field knowledge graph.This system can receive natural questions from users in the field of earthquake disaster prevention and control,analyze the intention of the question sentence through questionanswering algorithm,construct the query statement,search and return the answer in the knowledge graph,and present the knowledge network related to the question-answering in a visualized form.This paper has 41 figures,19 tables and 75 references.