摘要
传统方法采用数据挖掘的方式构建运维数据关联知识库,返回的关联信息并非用户想要的最终结果,不符合用户意图。为此,提出一种新的基于语义分析的运维数据关联知识库构建方法。给出运维数据关联知识库结构。采用文档主题生成模型LDA获取关键词,通过语义相似度求出知识案例与检索词间的相近程度,将和检索词相似度最高的结果作为检索结果。对运维知识项的各个属性赋予权值,依据构成元素的权值与知识库各个属性的匹配关系计算权值和,求出相似度。采用关键词权重量化法对检索结果进行处理。将构建知识库应用于动车系统中,结果表明,通过构建知识库可有效检索运维数据关联知识,为运维人员提供依据。
The traditional method uses data mining to construct the operation and maintenance data association knowledge base,and the related information returned is not the final result that the user wants,which is not in conformity with the user s intention.To this end,a new method of constructing operation and data association knowledge base based on semantic analysis is proposed.The operational knowledge database structure is given.The key words are obtained by using the document theme generation model LDA,the similarities between knowledge cases and search terms are obtained through the semantic similarity,and the result with the highest similarity to the search terms is taken as the search result.Weights and weights are assigned to all the attributes of the operation and maintenance knowledge item,and the similarity is calculated based on the matching between the weight of the constituent elements and each attribute of the knowledge base.The keyword weight quantification method is used to process the retrieval results.The construction knowledge base is applied to the motor car system.The results show that knowledge base can effectively retrieve the knowledge of operation and maintenance data and provide the basis for operation and maintenance personnel.
作者
来骥
马跃
吴舜
那琼澜
LAI Ji;MA Yue;WU Shun;NA Qiong-lan(State Grid Jibei Information&Telecommunication Company,Beijing 100053,China;Columbia University in the City of New York,United States of America;North China Electric Power University,Beijing 102206,China)
出处
《科学技术与工程》
北大核心
2018年第19期218-223,共6页
Science Technology and Engineering
关键词
语义分析
运维数据
关联
知识库
构建
semantic analysis
operation and maintenance data
association
knowledge base
construction
