节点文献

Web信息搜索技术的研究

【作者】 杨彬

【导师】 康慕宁;

【作者基本信息】 西北工业大学 , 计算机科学与技术, 2007, 硕士

【摘要】 随着万维网的发展,Web上的信息资源正在以前所未有的速度增长。面对海量的数据,用户常常无法从中找到自己所需要的数据。如何使用户能够在网络中快速,准确的找到所需要的数据是Web信息检索面临的挑战。 搜索引擎技术的出现,为用户提供了一种在Web中检索信息的简单的方法,使用户能够通过关键字进行相关资源的搜索。但是用户所需的资源种类不同,通用搜索引擎难以提供给用户足够的资源,因此出现了针对特定领域的搜索服务。RSS新闻搜索就是这类应用,它仅仅搜索RSS新闻资源。同时,越来越多的网络应用采用了B/S模式,因此出现了许多集成在浏览器上的搜索服务,并提供其他方便用户的附加功能。 本文首先介绍了信息检索技术的基本概念和模型,介绍了搜索引擎和元搜索引擎的基本结构;对基于链接分析的搜索引擎排序算法PageRank和HITS进行了分析和对比,在此基础上提出了基于概念的权重PageRank改进算法以及为页面标记概念的两种方法;提出了基于用户反馈的结果融合排名算法;详细介绍了RSS新闻搜索平台的结构,数据库模式设计,搜索操作的性能优化方法,主客观结合的新闻排名机制;最后介绍了一种浏览器插件,它主要提供一种为页面进行概念标记的方法,同时提供元搜索接口等其他服务。

【Abstract】 With the development of World Wide Web, the amount of data on web expands dramatically. Facing so many data on web, the users always lose their way to find the specific data. It is a challenge in web information retrieval to make the user find the data accurately and quickly.Search Engine provides a simple method to find webpage. Users can find the webpage by telling the Search Engine the keyword. Users always demand different types of resources, but general Search Engine cannot provide sufficient resource for users. Some specific search service emerges. For example, RSS search service’s aim is to only search the RSS news resources specifically. More and more web applications use Browser/Server architecture, so that many tools embedded in browsers has been implemented. These tools always provide an integrated search service and other function convenient for users.The paper gives an overview of basic concepts and modules of Information Retrieval, and the architecture of Search Engine and Meta Search Engine. Compares the PageRank and HITS algorithm, and then proposes a weighted PageRank algorithm based on concepts, and two methods of tagging concepts for webpage. Proposes a result merge ranking method based on users’ feedback. The paper describes the architecture of the RSS search platform in detail, including the schemas of the database, the optimization of search operation, and the ranking mechanism of RSS news. At last, describes a toolbar embedded in a browser, which provides a method of tagging concepts and meta serach services.

  • 【分类号】TP391.3
  • 【被引频次】2
  • 【下载频次】350
节点文献中: