搜索引擎的情报感知与刻画功能协同研究  

Collaborative research on intelligence perception and characterization of search engines

作　　者：张莹莹 刘秀磊[1] 白雪瑞 尹静 ZHANG Yingying;LIU Xiulei;BAI Xuerui;YIN Jing(Laboratory of Data and Scientific Intelligence Analysis,Beijing Information Science&Technology University,Beijing 100101,China;Key Laboratory of Frontier and Education Bureau,Chinese Academy of Sciences,Beijing 100864,China;Science and Technology Information Center,China Railway Research Institute Co.,Ltd,Chengdu 611731,China)

机构地区：[1]北京信息科技大学数据与科学情报分析实验室,北京100101 [2]中国科学院前沿与教育局重点实验室,北京100864 [3]中铁科学研究院有限公司科技情报中心,成都611731

出　　处：《北京信息科技大学学报（自然科学版）》2019年第6期19-24,共6页Journal of Beijing Information Science and Technology University

基　　金：国家自然科学基金资助项目(61601039);国家重点研发计划资助项目(2016YFC0801407);中国中铁2017年重点项目(41272321)

摘　　要：针对情报领域数据存在的收集、处理、感知、刻画困难等问题,提出了一种以知识图谱为刻画结果的搜索引擎,从而感知领域细微的变化态势;使用领域本体库进行知识抽取及挖掘,构建了领域知识图谱;采用基于知识图谱的分布式索引的动态构建、分布式倒排索引查询机制和智能排序算法等技术,构建了基于知识图谱的检索系统,改善了传统搜索引擎检索结果内容单一、召回率低的问题;实现了高效快速地提炼出可满足情报人员需要的、可理解的、可利用的资源,从而为企业决策者提供数据支持。To implement data collection, processing, perception and characterization in the field of intelligence, a search engine with knowledge graph as the result is proposed to perceive the subtle changes of the field. The domain ontology library is used for knowledge extraction and mining. The domain knowledge graph is used. The knowledge index-based dynamic index construction, distributed inverted index query mechanism and intelligent sorting algorithm are used to construct a knowledge graph-based retrieval system, which overcome the problems of single content and low recall rate in traditional search engine. It provides efficient and fast refinement of understandable and available resources that meet the needs of intelligence personnel, providing data support for business decision makers.

关 键 词：情报感知 情报刻画 知识图谱 动态索引 分布式检索 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]