基于知识图谱的学术信息检索系统研究

Research on Academic Information Retrieval System Based on Knowledge Graph

知识图谱理论已经应用在Google、百度、搜狗等搜索引擎,进入了知识检索新里程。然而科技论文检索领域并没有一套完整的以知识图谱理论为体系的科技论文检索系统。这样使得科技工作者在检索科技论文或作者的时候不能全方位地了解一篇论文或一位作者相关的论文或作者。本文首先综述了知识图谱及其目前的应用领域,设计了科技论文检索系统基于作者和论文的知识图谱构建模型。然后设计了学术信息实体抽取、知识融合等知识图谱构建方法,最后实验部分利用某出版行业现有的数据和百度学术搜索数据、ElasticSearch检索服务组件、Graphviz图形绘制工具构建了基于知识图谱的科技论文检索系统。

The theory of knowledge graph has been applied to search engines such as Google,Baidu,and Sogou,and has entered a new mileage of knowledge retrieval.However,there is not a complete set of retrieval system for scientific papers based on knowledge graph theory.This makes it impossible for a researcher to get a full understanding of information related to a paper or an author when searching for a scientific paper or author.This paper first reviews the knowledge graph and its current application area,and then designs a retrieval system for scientific papers,based on the knowledge graph of authors and papers.Followed by the design of the construction of academic information extraction,knowledge fusion of knowledge graph constructing method,finally the experimental part of the use of existing data in some publishing industry and data from Baidu scholar,elasticSearch search service components and graphviz graphics tools to construct a retrieval system for scientific papers based on knowledge graph.