面向智能搜索应用的水利知识图谱构建  

Construction of water conservancy knowledge graph for intelligent search application

作　　者：高凤宁 高祥涛[2] 曹帅[2] 朱向荣 司存友 胡伟[1] GAO Fengning;GAO Xiangtao;CAO Shuai;ZHU Xiangrong;SI Cunyou;HU Wei(Department of Computer Science and Technology,Nanjing University,Nanjing 210023,China;Jiangsu Hydrology and Water Resources Survey Bureau,Nanjing 210029,China)

机构地区：[1]南京大学计算机科学与技术系,江苏南京210023 [2]江苏省水文水资源勘测局,江苏南京210029

出　　处：《江苏水利》2021年第10期59-64,共6页Jiangsu Water Resources

基　　金：江苏省水利科技项目(2019046)。

摘　　要：在水利知识图谱的基础上,结合字符串相似度以及word2vec生成的词嵌入的余弦相似度,设计了关联属性的语义查询算法,实现了异构水利知识的融合。在网页端搭建了一个智能搜索应用,通过用户实验,验证了基于水利知识图谱的智能搜索应用可以降低水利领域从业人员的数据获取难度,增强对水利领域相关知识的理解。On the basis of water conservancy knowledge graph,combined with the string similarity and the cosine similarity of word embedding generated by word2vec,semantic query algorithm of associated attributes was designed to achieve the integration of heterogeneous water conservancy knowledge.After building an intelligent search application on the web,user experiments verified that the intelligent search application based on the water conservancy knowledge graph could reduce the difficulty of data acquisition for practitioners in the water conservancy field and enhance the understanding of relevant knowledge in the water conservancy field.

关 键 词：水利知识图谱 知识融合 词嵌入 语义搜索 

分 类 号：TM734[电气工程—电力系统及自动化]