本文選題：Web服務(wù)　切入點(diǎn)：用戶信息　出處：《南京大學(xué)》2014年碩士論文

【摘要】：作為一種基于互聯(lián)網(wǎng)標(biāo)準(zhǔn)和XML技術(shù)的新型分布式計(jì)算模型,Web服務(wù)在電子商務(wù)和企業(yè)應(yīng)用集成等分布式平臺(tái)上發(fā)揮越來越重要的作用。隨著互聯(lián)網(wǎng)上Web服務(wù)數(shù)量的指數(shù)型增長,如何主動(dòng)感知用戶需求、挖掘用戶個(gè)人偏好并為用戶提供最感興趣的服務(wù)選擇列表,已經(jīng)成為Web服務(wù)研究領(lǐng)域的熱點(diǎn)問題。目前Web服務(wù)推薦研究中最為常用的是協(xié)同過濾算法,包括基于用戶的協(xié)同過濾、基于服務(wù)的協(xié)同過濾以及兩者的結(jié)合。協(xié)同過濾算法中最容易出現(xiàn)的就是稀疏矩陣和冷啟動(dòng)問題,所以針對(duì)傳統(tǒng)的協(xié)同過濾算法中存在的不足,本文將社交網(wǎng)絡(luò)結(jié)合到Web服務(wù)推薦算法中,提出了一種基于用戶信任網(wǎng)絡(luò)和偏好的Web服務(wù)推薦算法,即首先提出基于用戶關(guān)系和偏好的服務(wù)推薦算法,在驗(yàn)證算法有效性的基礎(chǔ)上,通過深入挖掘用戶信息構(gòu)建信任網(wǎng)絡(luò),將用戶偏好算法和信任網(wǎng)絡(luò)相結(jié)合,為用戶提供更為有效的服務(wù)推薦。論文的主要貢獻(xiàn)如下：首先,針對(duì)Web服務(wù)QoS屬性的多樣性,提出了一種基于多QoS值的相似度計(jì)算方法,這種方法可以直接計(jì)算有多種QoS屬性的服務(wù)的相似性和訪問服務(wù)的用戶的相似性。在此基礎(chǔ)上,提出了一種基于用戶關(guān)系和偏好的Web服務(wù)推薦算法,通過使用服務(wù)信息對(duì)服務(wù)進(jìn)行聚類,將用戶-服務(wù)矩陣轉(zhuǎn)化為用戶-服務(wù)類矩陣來實(shí)現(xiàn)稀疏矩陣降維,并從社交網(wǎng)絡(luò)中獲取充分的用戶信息和用戶關(guān)系。通過挖掘用戶和服務(wù)類之間的關(guān)系,根據(jù)用戶偏好將用戶劃分為不同的興趣類并提取出每個(gè)類的顯著用戶特征,再結(jié)合社交網(wǎng)絡(luò)中的新用戶信息和興趣標(biāo)簽,通過與用戶興趣類的用戶特征和服務(wù)類標(biāo)簽進(jìn)行比對(duì),完成對(duì)新用戶的推薦,從而解決推薦系統(tǒng)的冷啟動(dòng)問題。其次,提出了一種根據(jù)用戶信息構(gòu)建用戶信任網(wǎng)絡(luò)模型的方法,可充分利用社交網(wǎng)絡(luò)中的用戶信息,深入挖掘用戶潛在關(guān)系。同時(shí),為了提高推薦的準(zhǔn)確性,使用主成分分析算法對(duì)用戶偏好算法進(jìn)行優(yōu)化,對(duì)同一興趣類中的用戶關(guān)系進(jìn)行更為嚴(yán)格的劃分,并將用戶信任網(wǎng)絡(luò)與之結(jié)合,構(gòu)成了基于用戶信任網(wǎng)絡(luò)和偏好的Web服務(wù)推薦算法。與用戶關(guān)系偏好算法相比,這種推薦算法在擴(kuò)充服務(wù)推薦范圍的基礎(chǔ)上又提高了服務(wù)篩選的標(biāo)準(zhǔn),既考慮了用戶偏好的相似性,又深入挖掘了用戶的潛在信任關(guān)系,可以為新用戶推薦滿足用戶需求且具有一定可信度的服務(wù)。再次,在工具實(shí)現(xiàn)和實(shí)驗(yàn)分析上,完成了Web服務(wù)推薦工具的開發(fā),并針對(duì)基于用戶關(guān)系和偏好、基于用戶信任網(wǎng)絡(luò)和偏好的服務(wù)推薦算法進(jìn)行了充分的實(shí)驗(yàn)。實(shí)驗(yàn)結(jié)果顯示,與傳統(tǒng)的基于協(xié)同過濾算法相比,文中提出的兩種推薦算法具有更高的推薦準(zhǔn)確率,尤其是基于用戶信任網(wǎng)絡(luò)和偏好的推薦算法,其推薦準(zhǔn)確率在基于用戶關(guān)系和偏好的推薦算法基礎(chǔ)上有明顯提高。
[Abstract]:As a new distributed computing model based on Internet standards and XML technology, web services play an increasingly important role in distributed platforms such as e-commerce and enterprise application integration.With the exponential growth of the number of Web services on the Internet, it has become a hot issue in the research field of Web services that how to actively perceive user needs, mine user preferences and provide users with the most interesting list of service choices.At present, collaborative filtering algorithms are the most commonly used in the research of Web services recommendation, including user-based collaborative filtering, service-based collaborative filtering and the combination of the two.The problem of sparse matrix and cold start is the most common problem in the collaborative filtering algorithm. Therefore, in view of the shortcomings of the traditional collaborative filtering algorithm, this paper combines the social network into the Web services recommendation algorithm.In this paper, a Web service recommendation algorithm based on user trust network and preference is proposed. Firstly, a service recommendation algorithm based on user relationship and preference is proposed. On the basis of verifying the validity of the algorithm, the trust network is constructed by mining user information deeply.The user preference algorithm and trust network are combined to provide more efficient service recommendation for users.The main contributions of this paper are as follows: firstly, a similarity calculation method based on multiple QoS values is proposed for the diversity of QoS attributes of Web services.This method can directly calculate the similarity of services with multiple QoS attributes and the similarity of users accessing services.On this basis, a Web service recommendation algorithm based on user relationship and preference is proposed. By using service information to cluster services, the user-service matrix is transformed into user-service class matrix to realize sparse matrix dimensionality reduction.And from the social network to obtain adequate user information and user relations.By mining the relationship between users and service classes, the users are divided into different interest classes according to their preferences, and the salient user characteristics of each class are extracted, and then the new user information and interest tags in social networks are combined.By comparing with the user characteristics of user interest class and the label of service class, the recommendation of new users is completed, and the cold start problem of recommendation system is solved.Secondly, a method of constructing user trust network model based on user information is proposed, which can make full use of user information in social network and tap the potential relationship of users.At the same time, in order to improve the accuracy of recommendation, the principal component analysis (PCA) algorithm is used to optimize the user preference algorithm, the user relationship in the same interest class is more strictly divided, and the user trust network is combined with it.A Web service recommendation algorithm based on user trust network and preference is constructed.Compared with the user relationship preference algorithm, this recommendation algorithm not only improves the standard of service selection on the basis of extending the range of service recommendation, but also takes into account the similarity of user preference and excavates the potential trust relationship of users.New users can recommend services that meet their needs and have a certain degree of credibility.Thirdly, in the aspect of tool implementation and experimental analysis, the development of Web service recommendation tool is completed, and a full experiment is carried out on the service recommendation algorithm based on user relationship and preference, based on user trust network and preference.The experimental results show that compared with the traditional collaborative filtering algorithm, the proposed two recommendation algorithms have higher recommendation accuracy, especially the recommendation algorithm based on user trust network and preference.The recommendation accuracy is improved obviously on the basis of the recommendation algorithm based on user relationship and preference.
【學(xué)位授予單位】：南京大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2014
【分類號(hào)】：TP391.3;TP393.09

【相似文獻(xiàn)】

相關(guān)期刊論文 前10條

1 徐義峰;徐云青;劉曉平;;一種基于時(shí)間序列性的推薦算法[J];計(jì)算機(jī)系統(tǒng)應(yīng)用;2006年10期

2 余小鵬;;一種基于多層關(guān)聯(lián)規(guī)則的推薦算法研究[J];計(jì)算機(jī)應(yīng)用;2007年06期

3 張海玉;劉志都;楊彩;賈松浩;;基于頁面聚類的推薦算法的改進(jìn)[J];計(jì)算機(jī)應(yīng)用與軟件;2008年09期

4 張立燕;;一種基于用戶事務(wù)模式的推薦算法[J];福建電腦;2009年03期

5 王晗;夏自謙;;基于蟻群算法和瀏覽路徑的推薦算法研究[J];中國科技信息;2009年07期

6 周珊丹;周興社;王海鵬;倪紅波;張桂英;苗強(qiáng);;智能博物館環(huán)境下的個(gè)性化推薦算法[J];計(jì)算機(jī)工程與應(yīng)用;2010年19期

7 王文;;個(gè)性化推薦算法研究[J];電腦知識(shí)與技術(shù);2010年16期

8 張愷;秦亮曦;寧朝波;李文閣;;改進(jìn)評(píng)價(jià)估計(jì)的混合推薦算法研究[J];微計(jì)算機(jī)信息;2010年36期

9 夏秀峰;代沁;叢麗暉;;用戶顯意識(shí)下的多重態(tài)度個(gè)性化推薦算法[J];計(jì)算機(jī)工程與應(yīng)用;2011年16期

10 楊博;趙鵬飛;;推薦算法綜述[J];山西大學(xué)學(xué)報(bào)(自然科學(xué)版);2011年03期

相關(guān)會(huì)議論文 前10條

1 王韜丞;羅喜軍;杜小勇;;基于層次的推薦:一種新的個(gè)性化推薦算法[A];第二十四屆中國數(shù)據(jù)庫學(xué)術(shù)會(huì)議論文集（技術(shù)報(bào)告篇）[C];2007年

2 唐燦;;基于模糊用戶心理模式的個(gè)性化推薦算法[A];2008年計(jì)算機(jī)應(yīng)用技術(shù)交流會(huì)論文集[C];2008年

3 秦國;杜小勇;;基于用戶層次信息的協(xié)同推薦算法[A];第二十一屆中國數(shù)據(jù)庫學(xué)術(shù)會(huì)議論文集（技術(shù)報(bào)告篇）[C];2004年

4 周玉妮;鄭會(huì)頌;;基于瀏覽路徑選擇的蟻群推薦算法:用于移動(dòng)商務(wù)個(gè)性化推薦系統(tǒng)[A];社會(huì)經(jīng)濟(jì)發(fā)展轉(zhuǎn)型與系統(tǒng)工程——中國系統(tǒng)工程學(xué)會(huì)第17屆學(xué)術(shù)年會(huì)論文集[C];2012年

5 蘇日啟;胡皓;汪秉宏;;基于網(wǎng)絡(luò)的含時(shí)推薦算法[A];第五屆全國復(fù)雜網(wǎng)絡(luò)學(xué)術(shù)會(huì)議論文（摘要）匯集[C];2009年

6 梁莘q，

本文編號(hào)：1728958