本文選題：信息檢索 + 智能代理��； 參考：《哈爾濱理工大學(xué)》2007年碩士論文

【摘要】： 隨著Internet的飛速發(fā)展,人們能夠比以往更容易、更直接地通過(guò)網(wǎng)絡(luò)獲取各種形式的信息�，F(xiàn)有的Internet搜索引擎如:Google、Yahoo、WebCrawler等,可以幫助人們搜索Internet上的各種信息。但由于語(yǔ)言的模糊性,詞語(yǔ)的多義性,利用現(xiàn)有搜索引擎用戶常常難以準(zhǔn)確地表達(dá)用戶興趣;而且不能區(qū)分用戶;他們也不能主動(dòng)從網(wǎng)絡(luò)上發(fā)現(xiàn)和收集用戶需要的信息,用戶要查詢同樣的興趣,只能再次搜索,己獲得最新的網(wǎng)頁(yè)內(nèi)容,浪費(fèi)了用戶大量的時(shí)間。 面對(duì)網(wǎng)絡(luò)信息服務(wù)的這種現(xiàn)狀,人們?cè)趯で笠环N將信息用戶感興趣的信息主動(dòng)推薦給用戶的服務(wù)方式,這便是個(gè)性化的主動(dòng)信息服務(wù)。在實(shí)現(xiàn)個(gè)性化的主動(dòng)信息服務(wù)中,智能Agent技術(shù)起到了至關(guān)重要的作用。 本課題針對(duì)目前信息檢索系統(tǒng)存在的不足,首先,在系統(tǒng)地介紹信息檢索研究現(xiàn)狀的基礎(chǔ)上對(duì)個(gè)性化信息檢索的發(fā)展、工作原理和現(xiàn)狀進(jìn)行了簡(jiǎn)要綜述,并對(duì)Agent技術(shù)做了介紹。然后,從現(xiàn)有問(wèn)題入手,開(kāi)發(fā)設(shè)計(jì)了一個(gè)基于Agent的個(gè)性化智能信息檢索系統(tǒng)模型。對(duì)基于Agent的個(gè)性化信息檢索系統(tǒng)的基本結(jié)構(gòu)、方法及相關(guān)技術(shù)進(jìn)行了研究。 該模型由用戶信息檢索個(gè)性Agent、信息搜索Agent和信息過(guò)濾Agent三個(gè)模塊構(gòu)成,分別對(duì)三個(gè)模塊中的關(guān)鍵技術(shù)進(jìn)行研究。信息檢索個(gè)性Agent研究是本文重點(diǎn)。用戶信息檢索個(gè)性Agent通過(guò)學(xué)習(xí)用戶的興趣,使其具有一定的智能性。通過(guò)用戶信息需求的表達(dá)和信息反饋,形成并訓(xùn)練用戶信息檢索個(gè)性模型。在對(duì)用戶個(gè)性化進(jìn)行深入研究時(shí),提出了一種改進(jìn)的用戶興趣模型,并詳細(xì)說(shuō)明了其生成和更新實(shí)現(xiàn)算法。再次,信息搜索Agent通過(guò)查詢代理與Internet搜索引擎連接,既可實(shí)現(xiàn)元搜索,又可以在返回的網(wǎng)址較少或不滿足用戶的要求時(shí),使用自身搜索工具在網(wǎng)絡(luò)上自主搜索,而且搜索算法從查詢代理返回的網(wǎng)址出發(fā)進(jìn)行搜索,減少了搜索的范圍,加快了搜索的速度。信息過(guò)濾Agent根據(jù)用戶已有的信息資源分析用戶喜好,采用向量空間法進(jìn)行信息過(guò)濾。接著本文對(duì)具體實(shí)現(xiàn)進(jìn)行了介紹,實(shí)現(xiàn)了系統(tǒng)的部分功能。 結(jié)果表明,該平臺(tái)可減少搜索范圍,加快搜索速度。最后,對(duì)本文的研究以及進(jìn)一步研究做了總結(jié)。
[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.
【學(xué)位授予單位】：哈爾濱理工大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2007
【分類號(hào)】：TP391.3

【引證文獻(xiàn)】

相關(guān)碩士學(xué)位論文 前3條

1 張維瑞;網(wǎng)絡(luò)招聘信息個(gè)性化推薦技術(shù)研究[D];大連海事大學(xué);2010年

2 李梅;改進(jìn)的K均值算法在中文文本聚類中的研究[D];安徽大學(xué);2010年

3 彭耶萍;基于WEB的智能化信息檢索系統(tǒng)研究[D];中南大學(xué);2009年

，

本文編號(hào)：1891130