發(fā)布時間：2018-05-16 07:34

  本文選題：社交網(wǎng)絡(luò) + 協(xié)同過濾　； 參考：《中國海洋大學》2014年碩士論文

【摘要】：社交網(wǎng)絡(luò)的迅速發(fā)展，給我們的生活帶來了各項便利，使得越來越多的用戶加入到其中來。社交網(wǎng)絡(luò)軟件也順應(yīng)潮流，源源不斷地被開發(fā)出來。在社交網(wǎng)絡(luò)中人們可以建立好友關(guān)系，通過好友關(guān)系查看好友的動態(tài)、日志等，也可以分享信息，實現(xiàn)即時通訊。社交網(wǎng)絡(luò)以交友為核心，逐漸滲透到各個領(lǐng)域中，成為人類社會交流的一個工具。人們也正享受著這種“足不出戶，能知天下事”的情景�，F(xiàn)在的智能手機更是把社交網(wǎng)絡(luò)展現(xiàn)的淋淋盡致，由于手機的普遍性和無線網(wǎng)絡(luò)的應(yīng)用，使得幾乎每個人手里都攥著這個社交網(wǎng)絡(luò)的載體。但是，面對如此多的用戶，如何準確、高效找到與自己志趣相投的用戶成了亟待解決的問題。好友推薦應(yīng)運而生，解決了這一難題。 本文在某政府內(nèi)網(wǎng)建立的微博應(yīng)用中設(shè)計實現(xiàn)了一種新的好友推薦方法，該方法整合政府所有人員的基本信息，政府人員發(fā)布的微博數(shù)目，以及應(yīng)用中用戶對推薦的反饋情況，實時的為用戶推薦好友。既解決了用戶繁多不知該添加誰的問題，又避免了新用戶加入的尷尬情況，擴大了用戶的交友圈。 本文的主要研究內(nèi)容如下： 首先，介紹了推薦系統(tǒng)的概念、流程以及相關(guān)的理論，列舉了幾種常見的推薦技術(shù)，對不同推薦技術(shù)的優(yōu)缺點也做了分析。 其次，詳細介紹了本論文使用的協(xié)同過濾技術(shù)，包括其核心思想、分類以及在不同分類下推薦算法的計算方法。并在此基礎(chǔ)上，設(shè)計了一種新的好友推薦方法，對此方法的功能和實現(xiàn)流程做了詳細設(shè)計。此方法的應(yīng)用解決了以往的推薦算法對新用戶，用戶反饋等問題的疏忽，為政府內(nèi)網(wǎng)微博的用戶提供了準確的好友推薦功能。 最后，將基于協(xié)同過濾技術(shù)的好友推薦方法用在已實現(xiàn)的政府內(nèi)網(wǎng)微博中。該系統(tǒng)的客戶端建立在當前最流行的Android系統(tǒng)上，實現(xiàn)了微博應(yīng)用的相關(guān)功能；由服務(wù)器為客戶端提供接口，主要實現(xiàn)數(shù)據(jù)存儲和好友推薦兩大功能模塊。 本文所實現(xiàn)的微博應(yīng)用，使公務(wù)員可以隨時隨地的進行網(wǎng)上交流學習互動，脫離辦公室電腦的束縛。系統(tǒng)除了實現(xiàn)現(xiàn)有的微博功能外，，還加入了好友推薦功能。該功能不僅僅依靠用戶的個人信息推薦，還依據(jù)用戶的微博使用以及用戶對推薦的反饋情況。這不僅為公務(wù)員的學習交流提供更大的便利，同時也擴充了用戶的使用量，更對用戶擴大社交圈和學習圈提供更好的服務(wù)。
[Abstract]:With the rapid development of social network, more and more users join our daily life. Social networking software has also been developed in a steady stream. In the social network, people can establish a good friend relationship, check the friends' dynamic, log and so on through the good friend relationship, also can share the information, realize instant communication. Social network, which is centered on making friends, has gradually penetrated into various fields and become a tool for human social communication. People are also enjoying the sight of staying in the house and knowing the world. Today's smartphones are all about social networking. Because of the popularity of mobile phones and the use of wireless networks, almost everyone has the carrier of the social network in their hands. However, in the face of so many users, how to find users with similar interests accurately and efficiently becomes an urgent problem. Good friend recommendation arises at the historic moment, solved this difficult problem. In this paper, a new friend recommendation method is designed and implemented in the Weibo application set up by a government intranet. This method integrates the basic information of all government personnel, the number of Weibo released by government personnel, and the feedback from users on the recommendation in the application. Recommend friends for users in real time. It not only solves the problem of many users do not know who to add, but also avoids the awkward situation of new users to join, and expands the circle of users to make friends. The main contents of this paper are as follows: Firstly, this paper introduces the concept, flow chart and related theories of recommendation system, enumerates several common recommendation technologies, and analyzes the advantages and disadvantages of different recommendation technologies. Secondly, the cooperative filtering technology used in this paper is introduced in detail, including its core idea, classification and the calculation method of recommendation algorithm under different classification. On this basis, a new friend recommendation method is designed, and the function and implementation flow of this method are designed in detail. The application of this method solves the neglect of the previous recommendation algorithms to the new users, user feedback and so on, and provides the accurate friend recommendation function for the Weibo users of the government intranet. Finally, the best friend recommendation method based on collaborative filtering technology is used in the implemented Weibo. The client of the system is based on the most popular Android system, which realizes the related functions of Weibo application, and the server provides the interface for the client, mainly realizes the two function modules of data storage and friend recommendation. The application of Weibo in this paper enables civil servants to communicate and learn online anytime and anywhere, and break away from the shackles of office computers. The system not only realizes the existing Weibo function, but also adds the good friend recommendation function. This function depends not only on the user's personal information recommendation, but also on the user's Weibo usage and the user's feedback on the recommendation. This not only provides greater convenience for the learning and communication of civil servants, but also expands the usage of users, and provides better services for users to expand their social circle and study circle.
【學位授予單位】：中國海洋大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：TP393.092

【參考文獻】

相關(guān)期刊論文 前10條

1 Bruce Antelman;李雯;;社交網(wǎng)絡(luò)[J];高校圖書館工作;2008年01期

2 歐陽為民;鄭誠;蔡慶生;;國際上關(guān)聯(lián)規(guī)則發(fā)現(xiàn)研究述評[J];計算機科學;1999年03期

3 嚴雋薇;黃勛;劉敏;朱延波;倪亥彬;;基于本體用戶興趣模型的個性化推薦算法[J];計算機集成制造系統(tǒng);2010年12期

4 曾小波;魏祖寬;金在弘;;協(xié)同過濾系統(tǒng)的矩陣稀疏性問題的研究[J];計算機應(yīng)用;2010年04期

5 張海燕,丁峰,姜麗紅;基于模糊聚類的協(xié)同過濾推薦方法[J];計算機仿真;2005年08期

6 慕春棣,tsinghua.edu.cn,戴劍彬,葉俊;用于數(shù)據(jù)挖掘的貝葉斯網(wǎng)絡(luò)[J];軟件學報;2000年05期

7 曾春,邢春曉,周立柱;基于內(nèi)容過濾的個性化搜索算法[J];軟件學報;2003年05期

8 李曉昀;陽小華;余穎;;基于隱性反饋分析的個性化推薦研究[J];計算機工程與設(shè)計;2009年16期

9 靳紅,楊艷紅;高校圖書館個性化信息服務(wù)研究綜述[J];現(xiàn)代情報;2004年06期

10 馬宏偉;張光衛(wèi);李鵬;;協(xié)同過濾推薦算法綜述[J];小型微型計算機系統(tǒng);2009年07期

本文編號：1896001