大数据背景下微博语义检索

发布时间：2018-05-12 12:43

本文选题：微博 + 语义检索　；参考：《情报杂志》2017年12期

【摘要】：[目的/意义]目前微博已经成为人们获取信息和发布信息的主要平台,现有微博平台的检索功能主要靠关键词匹配返回检索结果,使得检索结果无法满足用户需求。在大数据的背景下,实现微博语义检索已成为目前迫切需要解决的问题。[方法/过程]为了解决该问题,采用一种基于通用本体知识库的微博语义检索框架,利用通用本体知识库中概念之间的语义关系,借助概念相似度算法对用户查询进行扩展以及对微博短文本特征进行扩展与选择。[结果/结论]基于通用本体构建的微博语义检索框架,既可以从语义角度理解用户的查询并对其进行扩展,又可以克服微博短文本特征稀疏的缺点,从而可以提高微博检索的查全率和查准率,能有效解决大数据背景下传统微博检索不能满足用户检索需求的问题。最后对微博语义检索实现所涉及的相关问题进行了探讨。
[Abstract]:[purpose / significance] at present, Weibo has become the main platform for people to obtain and publish information. The retrieval function of the existing Weibo platform mainly depends on keyword matching to return the retrieval results, which makes the retrieval results unable to meet the needs of users. Under the background of big data, Weibo semantic retrieval has become an urgent problem to be solved. [method / process] in order to solve this problem, a Weibo semantic retrieval framework based on general ontology knowledge base is adopted, and the semantic relationship between concepts in general ontology knowledge base is utilized. With the help of conceptual similarity algorithm, the user query is extended and the Weibo short text feature is extended and selected. [results / conclusion] the Weibo semantic retrieval framework based on general ontology can not only understand and extend the query of users from the semantic point of view, but also overcome the shortcoming of sparse feature of Weibo short text. Therefore, the recall and precision of Weibo retrieval can be improved, and the problem that traditional Weibo retrieval can not meet the needs of users under the background of big data can be effectively solved. Finally, the related problems in the implementation of Weibo semantic retrieval are discussed.
【作者单位】：山西财经大学信息管理学院;中科院文献情报中心;
【基金】：山西省哲学社会科学“十二五”规划2015年度课题“山西智慧城市建设中多媒体大数据体系构建及应用研究”(编号:晋规办字【2015】3号) 山西省2016年度艺术科学规划课题“基于多技术融合的山西智慧博物馆体系构建研究”(编号:2016C18) 山西省教育科学“十三五”规划2016年度课题“‘互联网+’信息经济时代山西省大学生创新创业教育体系研究”(编号:GH-16053) 山西省研究生教育改革项目“慕课(MOOC)环境下研究生信息检索课程教学模式探索”(编号:2016JG64)研究成果之一
【分类号】：TP391.3

【相似文献】