【摘要】 针对冷鲜羊肉品质检测知识数据庞大、利用率低、结构复杂和知识碎片化等问题，构建了羊肉品质检测知识图谱；采用自顶向下和自底向上相结合的方法，围绕羊肉种类、检测技术和羊肉品质影响因素3个方面获取相关知识，解决知识碎片化问题。具体作法是：自顶向下构建知识模式层，通过本体建模形成知识图谱的概念框架；自底向上构建数据层，通过数据获取、知识抽取、融合、存储建立实体间关联关系。研究共抽取4 136个实体，18 104条三元组集合，并通过Neo4j数据库存储羊肉品质检测知识关联图数据，实现可视化表达。该研究的方法与结果将为羊肉品质检测知识智能检索与和智能问答系统构建提供参考与数据基础。更多还原

【Abstract】 To address the problems of huge data,low utilisation rate,complex structure and knowledge fragmentation of cold fresh lamb quality testing knowledge,this paper constructed a knowledge map for lamb quality testing; using a combination of top-down and bottom-up approaches,the knowledge was acquired around three aspects of lamb types, testing technologies and lamb quality influencing factors to solve the problem of knowledge fragmentation. The top-down knowledge model layer was constructed to form the conceptual framework of the knowledge map through ontology modelling,while the bottom-up data layer was constructed to establish inter-entity relationships through data acquisition,knowledge extraction,fusion and storage. In the study,a total of 4 136 entities and 18 104 triad sets were extracted,and the lamb quality inspection knowledge association graph data were stored through Neo4j database for visual representation. The methods and results of this study would provide a reference and data base for the construction of an intelligent retrieval and and intelligent question and answer system for lamb quality inspection knowledge.更多还原