Construction And Application Of Common Pig Disease Knowledge Graph Based On Graph Database Neo4j

Epidemic disease in pig farms is an important factor affecting the development of pig industry.But for now,pig disease prevention and control technology in the field of high-end talent training in research and speed lag behind the rapid expansion of pig breeding industry,become the bottleneck which restrict the development of the industry.In recent years,smart farming represented by Internet of things,artificial intelligence and other technologies has made certain progress,and has also accumulated a large number of common pig disease knowledge information for the animal husbandry and veterinary field.However,these data are widely distributed and very scattered and disorderly,and some data information is miscellaneous,which brings inconvenience to search.The pig disease question answering system based on knowledge graph is expected to solve this problem.The research content of this paper includes data acquisition and cleaning based on web crawler technology,knowledge graph theory,comparison of multiple knowledge extraction models and knowledge graph application,etc.In this study,the knowledge graph of common pig diseases in the process of pig breeding was constructed through web crawler,knowledge graph model,knowledge extraction,knowledge fusion,knowledge storage and other technologies.The wechat question answering robot based on the knowledge graph of common pig diseases was constructed by combining Python programming language,Neo4 j graph database and wechat platform,so as to realize an intelligent question answering system in the field of pig diseases.The main research contents and results are as follows:(1)A corpus of pig disease domain was constructed.According to the characteristics of pig disease data and the opinions of experts in the field of pig disease,the Chinese corpus of pig disease was determined,and the basic corpus of pig disease was obtained by web crawler technology.BIO annotation method and doccano annotation tool were used to annotate more than 50,000 words of porcine disease corpus,and then a porcine disease corpus was constructed.(2)A deep learn-based entity relationship extraction method for porcine disease information was proposed.The extraction effects of different knowledge extraction models on common pig disease text information were compared,and the extraction effects of Bert,XLNet and ERNIE pre-training models added on the basis of Bidirectional Long Short-Term Memory(Bi-LSTM)and Conditional Random Field(CRF)models were compared.Finally,the ERNIE-Bi-LSTM-CRF knowledge extraction model with the comprehensive evaluation index F1 value of 89.63% is adopted as the extraction model of this paper.(3)The knowledge graph of pig disease diagnosis is constructed.The extracted entities were fused and used as the data set for constructing the knowledge graph,and the entity data were stored in the Neo4 j graph database to realize the construction of the common pig disease knowledge graph.The knowledge graph integrates a large number of fragmented and heterogeneous pig disease information,which centralizes pig disease knowledge and is easy to manage and use.(4)An intelligent question answering system based on pig disease knowledge graph was constructed.On the basis of completing the construction of the common pig disease knowledge graph,a wechat pig disease question answering robot based on wechaty open source library was designed and implemented.The pig disease question answering robot classifies the questions queried by users through problem recognition,and constructs Cypher query questions for query in the Neo4 j graph database.Then,the wechat question answering robot automatically answers the questions raised by users,and users can query12 kinds of common pig diseases through the wechat chat interface.The fusion of knowledge graph technology and common pig disease knowledge in the process of breeding and production management can effectively improve the efficiency of retrieving pig disease knowledge,put forward a new scheme for the storage and query of common pig disease knowledge,and provide a new idea for the research of pig disease prevention technology and personnel training.At the same time,on this basis,the wechat question answering robot reduces the difficulty of querying pig disease knowledge,makes it more convenient for users to query common pig disease knowledge,and assists farmers to make production decisions.