【作者】 赵静；

【导师】 薛大伸；

【作者基本信息】 大连海事大学 ， 管理科学与工程， 2009， 硕士

【摘要】 随着网络技术的迅猛发展,WWW已成为信息发布、交互及获取的主要工具,它涉及新闻、广告、消费、金融、教育、电子商务等许多领域。人们要从这些海量的数据中查找数据和信息,使用最多的就是搜索引擎技术。本文针对MBA教育网络信息化的需求和特点,从查询结果个性化的角度出发,对教学资源的个性化搜索引擎进行了一个特色规划。随着用户群和教学资源库的不断壮大,如何人性化的理解用户的查询需求,尽可能准确地返回查询请求内涵和外延的结果;如何通过分析和研究个性化搜索引擎,掌握用户资料后进行分析,在用户搜索新的关键词时,能返回更有针对性的搜索结果,从而提高用户体验,是现有条件下搜索引擎研究应该努力着手解决的问题。本文研究了Web挖掘在教学资源搜索引擎中的应用,Web挖掘分为Web内容挖掘,Web使用挖掘和Web结构挖掘。论文主要对其中的前两项内容进行了研究,结构挖掘是从WWW的组织结构、Web文档结构和链接关系中推导知识。就搜索引擎技术领域来说,可以通过分析一个网页或整个网站链接和被链接的数量、对象,建立Web自身的链接结构模式,通过分析和研究基于链接结构的搜索结果排名算法,可以指导网站链接结构优化,有组织,有规划地提高网页在搜索结果中的排名,避免盲目处理造成的混乱结果。本文主要针对目前主流的PageRank算法,集中研究了该算法的计算方法、网页链接结构对PageRank值的影响,并分析该算法在独立网站、包含入站链接和出站链接等几种模型下的效果,提出了个性化PageRank算法优化策略。通过总结PageRank存在的一些问题,针对其中的主题漂移现象给出了改进后的个性化PageRank算法,并对其进行了验证。最后主要针对MBA教育资源搜索个性化引擎系统进行系统设计。更多还原

【Abstract】 With the rapid development of network technology,WWW has become the information,interactive and access tools.It relates to news,advertising,consumer, finance,education,e-commerce and many other fields.People who search data from these massive data and information frequently use search engine technology.In this paper,the network information needs and characteristics of MBA education,from the point of view personalized search results,teaching resources on the personalized search engines have been a feature of planning.As the user base and teaching resource library continue to grow,how to understand the human needs of the user’s query,as accurately as possible to return the connotation and extension of query results.Through analysis and research personalized search engine,after master user information we analysis. After new users search new key word,we can return more targeted search results, thereby enhancing the user experience,which is that search engine research should be made to address the problem under the existing conditions.This article also examined the teaching resources search engine in the Web Mining Application.Web mining is divided into Web content mining,Web usage mining and Web structure mining.This article mainly examined contents of the first two studies. Structure mining derive knowledge from WWW organizational structure,Web document structure and link relationship.On the field of search engine technology,it set up its own Web link structure model through the analysis of the number and targets of a web page or the entire web link and linked,can guide the web link structural optimization,improve organized and planly the page rank in the search results,avoid the confusion results by blindly the results of treatment through analysis and researching through ranking algorithms in search results based on the link structure.In this paper,the mainstream view is the current PageRank algorithm,the paper focuses on the influence of calculation method and the web link structure of the PageRank values.It analysis the algorithm effect in an independent website,including inbound links and outbound links,such as several models.It also poses optimization PageRank algorithm strategy.By summing up some problems of PageRank,it poses the improved optimization PageRank algorithm against the drift phenomenon and its verified.Finally its major task is that MBA educational resource personalization search engine system design.更多还原