Construction of Knowledge Graph of Animal Husbandry Based on Deep Learning
Aiming at the problems of highly specialized knowledge and difficult sharing in the field of animal husbandry,a knowledge graph of animal husbandry involving animal husbandry species,veterinary diseases,and veterinary drugs is constructed through four steps of data acquisition,ontology construction,knowledge extraction,and knowledge storage,which reduces the ap-plication threshold.Firstly,based on the 33 types of livestock breeds in the National Inventory of Livestock and Poultry Genetic Re-sources,animal husbandry species,veterinary diseases are collected,veterinary drugs from data sources such as the census infor-mation system of livestock and poultry genetic resources,the national veterinary drug basic database,and veterinary monographs.Secondly,the conceptual architecture of animal husbandry knowledge is defined,and the domain ontology of animal husbandry is constructed.Then deep learning-based methods and rule-based methods are used to extract entities and relationships in semi-struc-tured data and unstructured data,with a total of 6 138 entities and 27 870 triples.Finally,the extracted knowledge graph triplet da-ta is saved to the Neo4j graph database,which provides knowledge base support for subsequent applications such as intelligent medi-cal care and intelligent question answering.
国家科技期刊平台
NSTL
万方数据
