基于Agent的个性化智能信息检索系统

发布时间：2018-05-15 05:15

本文选题：信息检索 + 智能代理　；参考：《哈尔滨理工大学》2007年硕士论文

【摘要】： 随着Internet的飞速发展,人们能够比以往更容易、更直接地通过网络获取各种形式的信息。现有的Internet搜索引擎如:Google、Yahoo、WebCrawler等,可以帮助人们搜索Internet上的各种信息。但由于语言的模糊性,词语的多义性,利用现有搜索引擎用户常常难以准确地表达用户兴趣;而且不能区分用户;他们也不能主动从网络上发现和收集用户需要的信息,用户要查询同样的兴趣,只能再次搜索,己获得最新的网页内容,浪费了用户大量的时间。面对网络信息服务的这种现状,人们在寻求一种将信息用户感兴趣的信息主动推荐给用户的服务方式,这便是个性化的主动信息服务。在实现个性化的主动信息服务中,智能Agent技术起到了至关重要的作用。本课题针对目前信息检索系统存在的不足,首先,在系统地介绍信息检索研究现状的基础上对个性化信息检索的发展、工作原理和现状进行了简要综述,并对Agent技术做了介绍。然后,从现有问题入手,开发设计了一个基于Agent的个性化智能信息检索系统模型。对基于Agent的个性化信息检索系统的基本结构、方法及相关技术进行了研究。该模型由用户信息检索个性Agent、信息搜索Agent和信息过滤Agent三个模块构成,分别对三个模块中的关键技术进行研究。信息检索个性Agent研究是本文重点。用户信息检索个性Agent通过学习用户的兴趣,使其具有一定的智能性。通过用户信息需求的表达和信息反馈,形成并训练用户信息检索个性模型。在对用户个性化进行深入研究时,提出了一种改进的用户兴趣模型,并详细说明了其生成和更新实现算法。再次,信息搜索Agent通过查询代理与Internet搜索引擎连接,既可实现元搜索,又可以在返回的网址较少或不满足用户的要求时,使用自身搜索工具在网络上自主搜索,而且搜索算法从查询代理返回的网址出发进行搜索,减少了搜索的范围,加快了搜索的速度。信息过滤Agent根据用户已有的信息资源分析用户喜好,采用向量空间法进行信息过滤。接着本文对具体实现进行了介绍,实现了系统的部分功能。结果表明,该平台可减少搜索范围,加快搜索速度。最后,对本文的研究以及进一步研究做了总结。
[Abstract]:With the rapid development of Internet, people can obtain all kinds of information more directly and easily than ever before. Existing Internet search engines such as Google: Yahoo! Web Crawler can help people search for all kinds of information on Internet. However, because of the fuzziness of the language and the ambiguity of the words, it is often difficult to express the user's interest accurately by using the existing search engine users, and they can not distinguish the users, nor can they actively discover and collect the information that the users need from the network. Users can only search for the same interest again. They have to get the latest web content and waste a lot of time. In the face of the present situation of network information service, people are looking for a kind of service way that can actively recommend the information interested by information users, which is called personalized active information service. Intelligent Agent technology plays an important role in the realization of personalized active information service. This paper aims at the deficiency of the information retrieval system at present. Firstly, the development, working principle and present situation of personalized information retrieval are summarized on the basis of systematically introducing the present situation of information retrieval research, and the Agent technology is also introduced. Then, a personalized intelligent information retrieval system model based on Agent is developed from the existing problems. The basic structure, methods and related technologies of personalized information retrieval system based on Agent are studied. The model is composed of three modules: user Information Retrieval Personality Agent Information search Agent and Information filtering Agent. The key technologies of the three modules are studied respectively. The research of information retrieval personality Agent is the focus of this paper. User Information Retrieval Personality (Agent) makes it intelligent by learning user's interest. The user information retrieval personality model is formed and trained through the expression and feedback of user information requirements. In this paper, an improved user interest model is proposed, and its algorithm of generating and updating is described in detail. Thirdly, the information search Agent can connect with the Internet search engine through the query agent, which can not only realize the meta-search, but also use its own search tools to search the network independently when there are fewer URLs returned or do not meet the requirements of the users. Moreover, the search algorithm starts from the URL returned by the query agent, which reduces the scope of search and speeds up the search. Information filtering Agent analyzes user preferences according to existing information resources and adopts vector space method to filter information. Then the paper introduces the implementation and realizes some functions of the system. The results show that the platform can reduce the search range and speed up the search. Finally, the research and further research are summarized.
【学位授予单位】：哈尔滨理工大学
【学位级别】：硕士
【学位授予年份】：2007
【分类号】：TP391.3

【引证文献】