一种多视角新闻信息快速检索方法  

Method of fast news information retrieval from multiple perspective

作　　者：张志强[1,2] 张丽 周晓清[1,2] 王伟钧[2] 黎忠文[1,2] ZHANG Zhiqiang;ZHANG Li;ZHOU Xiaoqing;WANG Weijun;LI Zhongwen(Key Laboratory of Pattern Recognition and Intelligent Information Processing of Sichuan,Chengdu University,Chengdu 610106,China;College of Computer Science,Chengdu University,Chengdu 610106,China;School of Foreign Languages,Sichuan Normal University,Chengdu 610066,China)

机构地区：[1]成都大学模式识别与智能信息处理四川省高校重点实验室,四川成都610106 [2]成都大学计算机学院,四川成都610106 [3]四川师范大学外国语学院,四川成都610066

出　　处：《现代电子技术》2024年第18期57-64,共8页Modern Electronics Technique

基　　金：四川省教育厅教育科研课题(SCJG21A123);四川省高等学校人文社会科学重点研究基地-新建院校改革与发展研究中心项目(XJYX2023C03);全国高等院校计算机基础教育研究会-计算机基础教育教学研究项目(2023-AFCEC-149)。

摘　　要：信息检索一直是数据挖掘研究的重要领域,也是情报搜集的重要手段。如果采用人工手动搜索方式在数据库中检索信息,检索效率较低。为此,以新闻信息搜集为应用背景,以提高信息检索效率为目的,提出了一种多视角新闻信息快速检索方法。该方法以用户需要检索的多视角新闻描述信息为检索依据,利用关键词权重分割法处理多视角新闻描述信息来构建词汇向量,以词汇向量在新闻信息库中的检索结果为依据来构建词汇检索向量,以向量相似度计算模型为信息筛选规则,在新闻信息库中快速检索并筛选出用户需要搜集的新闻信息。文中给出了所提方法的设计模型、涉及的关键技术实现策略、数据处理流程以及实验对比分析。实验结果表明,该方法能够根据用户指定的多视角新闻描述信息,在新闻信息库中快速高效地检索并筛选出符合用户需求的新闻文章信息。Information retrieval has always been an important field of data mining research and an important mean of intelligence gathering.If the manual search method is used to retrieve information in the database,the retrieval efficiency is low.Therefore,a method of fast news information retrieval from multiple perspectives is proposed with the application background of news information collection and the purpose of improving the efficiency of the information retrieval.In this method,the multi-perspective news description information that users need to retrieve is used as the retrieval basis,and the keyword weight segmentation method is used to process multi-perspective news description information to build vocabulary vectors.The retrieval results of vocabulary vectors in the news information database are used as the basis to build vocabulary retrieval vectors.The vector similarity calculation model is used as the information screening rule to rapidly retrieve and screen out the news information that users need to collect in the news information database.The design model of the proposed method,the implementation strategy for key technologies involved in the method,data processing flow,and the comparative analysis of experiments are provided.The experimental results show that,according to multi-perspective news description information specified by users,this method can quickly and efficiently retrieve and screen out news article information that meets users′needs in the news information database.

关 键 词：多视角 新闻信息 信息检索 快速检索方法 词汇向量 词汇检索向量 向量相似度计算 

分 类 号：TN919-34[电子电信—通信与信息系统] TP391.1[电子电信—信息与通信工程]