基于二次排序的个性化中文搜索引擎设计与实现

发布时间：2018-07-01 09:20

本文选题：个性化搜索引擎 + 用户兴趣　；参考：《广西大学》2013年硕士论文

【摘要】：随着计算机系统性能的提高和网络技术的飞速发展,Internet成为全球最大的信息资源库,用户面临着“信息过载”的问题。如何为如此庞大的信息资源提供高效的信息检索服务,帮助用户在数据的海洋中快速找到需要的信息是搜索引擎急需解决的问题。解决上述问题的方法之一是研发个性化搜索引擎。如何实现个性化服务,已成为了搜索引擎技术研究中的热点之一。本文主要介绍了个性搜索引擎的研究现状以及现有搜索引擎的不足,结合第一和第二代搜索引擎技术,通过分析用户在使用搜索引擎时打开的链接Web页面,提出了一种基于由客户端自行即时分析、自行即时应答反馈、自行即时再排序的个性化搜索引擎模型——基于二次排序的个性化搜索引擎系统模型,并完成了一个基于二次排序个性化中文搜索引擎系统原型Myso搜索引擎。本论文所作的主要工作如下： (1)在研究个性化搜索引擎工作原理的基础上,设计和实现了基于二次排序的个性化搜索引擎系统模型-Myso搜索引擎。 (2)借鉴词频位置加权排序法,设计了基于中文分词系统ICTCIAS中词的eWeight计算网页Rank的公式与算法,计算网页初始的序列并简化Myso搜索引擎,。 (3)为了获取用户兴趣,利用隐式收集用户兴趣技术,设计了一种跟踪用户的行为的即时更新用户兴趣的方法；在此基础上,结合网页分类技术,提出了一种个性化排序方法一基于二次排序的个性化排序方法。
[Abstract]:With the improvement of computer system performance and the rapid development of network technology, Internet has become the largest information resource bank in the world, and users are facing the problem of "information overload". How to provide an efficient information retrieval service for such a huge information resource and help users quickly find the information needed in the ocean of data is an urgent problem for search engines to solve. One of the ways to solve these problems is to develop personalized search engines. How to realize personalized service has become one of the hotspots in the research of search engine technology. This paper mainly introduces the research status of personality search engine and the deficiency of the existing search engine. Combining with the first and second generation search engine technology, this paper analyzes the linked Web page that the user opens when using the search engine. This paper presents a personalized search engine model based on real-time analysis, feedback and reordering by the client, which is based on the second order of the personalized search engine system. And completed a personalized Chinese search engine system based on secondary sorting Myso search engine prototype. The main work of this paper is as follows: (1) on the basis of studying the working principle of personalized search engine, This paper designs and implements the personalized search engine model-Myso search engine based on quadratic ordering. (2) the formula and algorithm of eWeight computing web page Rank based on Chinese word segmentation system ICTCIAS is designed. The initial sequence of web pages is calculated and the Myso search engine is simplified. (3) in order to gain users' interest, an instant updating method of user interest is designed to track the user's behavior by using the technology of gathering user's interest implicitly. Combined with the technology of web page classification, a personalized sorting method based on secondary sorting is proposed.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.3

【参考文献】