【摘要】： 網(wǎng)絡(luò)技術(shù)的發(fā)展帶來(lái)了可獲取信息資源的極大豐富，但是網(wǎng)絡(luò)資源的無(wú)序、良莠不齊等缺點(diǎn)也給用戶獲取網(wǎng)絡(luò)信息帶來(lái)了困難。網(wǎng)絡(luò)信息獲取服務(wù)是指在互聯(lián)網(wǎng)上，針對(duì)個(gè)人用戶的網(wǎng)絡(luò)信息需求，以現(xiàn)代信息技術(shù)為手段，向用戶提供所需的互聯(lián)網(wǎng)信息產(chǎn)品及服務(wù)，其服務(wù)模式包括信息拉取和信息推送。自適應(yīng)網(wǎng)絡(luò)信息獲取服務(wù)技術(shù)，則是根據(jù)用戶需求、信息源特征、系統(tǒng)負(fù)載等因素，自適應(yīng)地動(dòng)態(tài)調(diào)整自身行為，高效、人性化地提供高質(zhì)量的信息。 準(zhǔn)確、全面地感知用戶需求，是實(shí)現(xiàn)網(wǎng)絡(luò)信息獲取服務(wù)的基礎(chǔ)。網(wǎng)絡(luò)用戶既是網(wǎng)絡(luò)信息資源的利用者又是提供者，因此可以通過(guò)分析用戶的瀏覽內(nèi)容、行為和發(fā)布的信息等來(lái)獲取用戶需求。獲得用戶需求后，如何在浩瀚的網(wǎng)絡(luò)信息資源中篩選出相關(guān)的信息，并以更人性化的方式展現(xiàn)給用戶，是網(wǎng)絡(luò)信息獲取服務(wù)成功的關(guān)鍵。此外，用戶對(duì)信息獲取的時(shí)效性通常有較高的要求，如何保障信息獲取系統(tǒng)的性能也是網(wǎng)絡(luò)信息獲取服務(wù)的重要研究?jī)?nèi)容之一。 為解決上述問(wèn)題，本文首先提出了一種基于查詢歧義性衡量的自適應(yīng)信息拉取技術(shù)。對(duì)用戶請(qǐng)求進(jìn)行歧義性衡量，根據(jù)其歧義性自適應(yīng)地決定結(jié)果的展現(xiàn)方式；在結(jié)果篩選和展現(xiàn)方面，分別提出了多特征融合排序算法和聚類算法；并在互聯(lián)網(wǎng)頗具代表的新興資源：多媒體信息(以圖像為例)和更新頻繁的動(dòng)態(tài)資源(以博客為例)上得到了驗(yàn)證。 其次，本文針對(duì)網(wǎng)絡(luò)活動(dòng)中的信息發(fā)布者和信息瀏覽者各提出了一種基于個(gè)性化建模的自適應(yīng)信息推送技術(shù)：對(duì)于信息發(fā)布者，以當(dāng)前網(wǎng)絡(luò)流行的博客這一個(gè)性化信息發(fā)布平臺(tái)為研究環(huán)境，提出了一種利用博客文章對(duì)用戶進(jìn)行長(zhǎng)短期興趣建模的方法，并對(duì)博客空間進(jìn)行社群劃分，實(shí)現(xiàn)了興趣相似好友的推薦；對(duì)于信息瀏覽者，利用用戶當(dāng)前瀏覽網(wǎng)頁(yè)的內(nèi)容作為用戶個(gè)性信息的表征，提出了一種基于情感和主題分析的上下文廣告推薦技術(shù)，使推送的廣告不僅主題相關(guān)，而且與網(wǎng)頁(yè)內(nèi)容中潛在的用戶情感相符合，，從而更具針對(duì)性。 接著，針對(duì)網(wǎng)絡(luò)信息獲取服務(wù)在性能、可擴(kuò)展性等方面的需求，以信息拉取服務(wù)的典型應(yīng)用——搜索引擎為切入點(diǎn)，提出了一種具有較好可擴(kuò)展性的混合型分布式索引組織策略(Loc-Glob)。并在Loc-Glob索引組織策略之上進(jìn)行性能優(yōu)化：基于索引詞負(fù)載及動(dòng)態(tài)變化查詢流，重新分布和冗余索引；基于索引服務(wù)器的實(shí)時(shí)系統(tǒng)負(fù)載，實(shí)現(xiàn)查詢路徑的自適應(yīng)優(yōu)化。 基于上述研究，本文設(shè)計(jì)并實(shí)現(xiàn)了一個(gè)采用自適應(yīng)技術(shù)的博客空間信息獲取原型系統(tǒng)，提供了博客搜索引擎、博客好友推薦、廣告推薦等多種應(yīng)用服務(wù)，驗(yàn)證了本文針對(duì)信息拉取和信息推送兩類服務(wù)模式提出的多項(xiàng)自適應(yīng)技術(shù)的可行性。 文章最后對(duì)本文的研究工作進(jìn)行了總結(jié)和展望。
[Abstract]:The development of network technology has brought about great richness of available information resources, but the disorder and uneven quality of network resources have also brought difficulties to users in obtaining network information. The service mode of Internet information products and services is information pulling and information pushing. Adaptive network information acquisition service technology adaptively adjusts its behavior dynamically according to user needs, information source characteristics, system load and other factors, and provides high-quality information efficiently and humanely.
Accurate and comprehensive perception of users'needs is the basis of network information acquisition services. Network users are both users and providers of network information resources. Therefore, users' needs can be obtained by analyzing users'browsing content, behavior and published information. Screening out relevant information and presenting it to users in a more humane way is the key to the success of network information acquisition service. In addition, users usually have higher requirements for the timeliness of information acquisition. How to ensure the performance of information acquisition system is also one of the important research contents of network information acquisition service.
In order to solve the above problems, an adaptive information pull-out technique based on query ambiguity measurement is proposed in this paper. It has been validated in the emerging resources of the Internet: multimedia information (for example, images) and frequently updated dynamic resources (for example, blogs).
Secondly, this paper proposes an adaptive information push technology based on personalized modeling for information publishers and information browsers in network activities. For information publishers, a personalized information publishing platform, blog, which is popular on the Internet, is taken as the research environment, and a blog article is proposed to make use of the length of users. Interest modeling method is used to divide the blog space into communities, and the recommendation of friends with similar interests is realized. For information browsers, the content of the user's current browsing web page is used as the representation of the user's personality information, and a context advertising recommendation technology based on emotion and topic analysis is proposed to make the pushed advertisement not only the theme. Relevance, and in line with potential user emotions in web content, is more targeted.
Then, aiming at the performance and scalability requirements of network information retrieval service, a hybrid distributed index organization strategy (Loc-Glob) with good scalability is proposed, which is based on the typical application of information pull-out service-search engine. Indexed word load and dynamic query flow are used to redistribute and redundant index, and real-time system load based on index server is used to optimize query path.
Based on the above research, this paper designs and implements a blog spatial information acquisition prototype system using adaptive technology, which provides a variety of application services such as blog search engine, blog friend recommendation, advertising recommendation and so on. It verifies the feasibility of the multi-adaptive technology proposed in this paper for the two types of service modes of information pull and information push.
At the end of the paper, the research work in this paper is summarized and prospected.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2008
【分類號(hào)】：G250.73

【引證文獻(xiàn)】

相關(guān)博士學(xué)位論文 前1條

1 王興中;鐵路調(diào)度指揮系統(tǒng)中數(shù)據(jù)存儲(chǔ)體系及索引結(jié)構(gòu)的研究[D];中國(guó)鐵道科學(xué)研究院;2011年

本文編號(hào)：2233080