一种基于用户标记的搜索结果排序算法

随着计算机网络的快速发展,网络上的信息量也日益纷繁复杂.如何准确、快速地帮助人们从海量网络数据中获取所需信息,这是目前搜索引擎首要解决的问题,为此,各种搜索排序算法应运而生.但是目前,网页信息的表达形式都十分简单,用户描述查询的形式更是十分简单,这就造成了在判断网页内容与用户查询相关性时十分困难.首先对现有的搜索引擎排序算法进行了分类总结,分析它们的优缺点.然后提出了一种基于用户反馈的语义标记的新方法,最后采用多种评估方法与Google搜索结果进行对比分析.实验结果表明,利用该方法所得到的排序结果比Google的排序结果更接近用户需求.

An User-Label Based Search Result Ranking Algorithm

With the development of computer network,the amount of lnformation on the network has become more and more every day.How to help people to find the information they needed exactly and quickly is the most important problem that search engines need to resolve immediately.As a result,a variety of ranking algorithms for search engine have appeared.However,the expression format of these Web pages is very simple and also the expression format of user query,which result in the difficulty of determining the relevance between user queries and Web page content.Firstly,the existing ranking algorithms of search engine are classified,and their advantage and disadvantage are analyzed.Then a new approach based on user feedback and semantic tags is proposed.Finally,the OSearch is compared with Google in several evaluation methods.The experimental results show that Osearch algorithm is closer to meet user needs than Google.