基于查询时间属性的海量博客数据检索方法  

Research on Blog Data Retrieval Method Based on Query Time Attributes

作　　者：曹纪清[1] CAO Ji- qing(Department of Information Technology,Suzhou Industrial Park Institute of Services Outsourcing,Suzhou 215123,Jiangsu,China)

机构地区：[1]苏州工业园区服务外包职业学院信息技术系,江苏苏州215123

出　　处：《内蒙古师范大学学报（自然科学汉文版）》2018年第3期248-252,258,共6页Journal of Inner Mongolia Normal University(Natural Science Edition)

基　　金：江苏省教育厅社科基金资助项目(2017SJB1953)

摘　　要：针对博客非正式、口语化语言多,以及输入查询的时间属性少等问题,提出一种利用查询的时间属性检索相关博客的高效方法.首先,自动计算每个查询在不同的时间间隔中的时间敏感度;然后,为查询建立一个时域剖面,并使用该剖面计算每个相关博文的时间得分;最后,基于博客与输入查询之间的时间和内容上的相关度,对所有的博客进行排名.使用irBlogs数据集对提出的方法进行了实验分析和比较,实验结果表明,所提方法在周期性时间敏感查询和事件敏感查询上表现出了较好的性能,优于其他博客检索方法.Aiming at the problems of informal blogs,colloquial languages and less time attributes of input query,an efficient method of retrieval relevant blogs by using the temporal attributes of queries is proposed.Firstly,the time sensitivity of each query in different time intervals is calculated automatically.Then,a time domain profile is established for the query,and the time score of each relevant blog is calculated with this profile.Finally,all blogs are ranked based on the relevance of time and content between the blog and the input query.Experimental analysis and comparison of the proposed method are carried out by using irBlogs dataset.The experimental results demonstrate that the proposed method shows performance on periodic time-sensitive queries and event-sensitive queries,which are superior to other blog retrieval methods.

关 键 词：博客检索 博客相关度 时间属性 时间敏感度 剖面 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]