基于新浪微博的好友推荐系统

发布时间：2018-06-20 02:09

本文选题：新浪微博 + 好友推荐　；参考：《燕山大学》2014年硕士论文

【摘要】：随着信息技术飞速发展及互联网的迅速普及，微博以其新颖的形式走入网民的视线，并迅速走红。2009年新浪推出“新浪微博”内测版，微博正式进入中文上网主流人群视野，并迅速发展，在国内引发了信息传播的“微革命”。随着微博在网民中日益火热，好友推荐成为了用户关注的焦点，同时也是本文研究的重要内容。首先，分析好友推荐算法的研究现状、核心思想以及目前存在的问题与不足。在此基础上，设计了基于用户的协同过滤好友推荐算法。该算法主要包括兴趣爱好、住址、二度人脉、教育工作信息和综合因素五个部分。其次，为前四个部分建立相似关系，通过关系算法公式分别计算四个部分的相似度得分，再为每个部分设置不同的权重。综合四部分的相似度得分计算出每个推荐结果的得分，根据分数筛选，，将综合相似度最高的前几项及其相似度分别显示给用户。最后，设计了新浪微博好友推荐系统，完成了对系统总体功能的设计。通过对注册登录、个人主页、个人资料、个人微博、个人相册、关注粉丝、分组管理及好友主页等功能模块设计相关数据库，实现了新浪微博好友推荐系统，并通过用户数据来验证了好友推荐算法的有效性，从而实现了快速帮助用户筛选有效信息的目标。
[Abstract]:With the rapid development of information technology and the rapid popularization of the Internet, Weibo has come to the attention of Internet users in its novel form, and has quickly become popular. In 2009, Sina launched the "Sina Weibo" internal version, and Weibo formally entered the mainstream Chinese Internet view. And rapid development, in the domestic triggered the information dissemination of the "micro-revolution." With the increasing popularity of Weibo among Internet users, friend recommendation has become the focus of user attention, but also an important part of this paper. Firstly, the research status, core idea, problems and shortcomings of friend recommendation algorithm are analyzed. On this basis, a collaborative filtering algorithm for friend recommendation based on user is designed. The algorithm mainly includes five parts: interest, address, second degree network, educational work information and comprehensive factors. Secondly, the similarity relation is established for the first four parts, the similarity score of the four parts is calculated by the formula of the relational algorithm, and the weight of each part is set up. By synthesizing the similarity scores of the four parts, the score of each recommendation result is calculated. According to the score selection, the first items with the highest synthetic similarity and their similarity are displayed to the users respectively. Finally, the Sina Weibo friend recommendation system is designed, and the overall function of the system is designed. By designing the relevant database for the function modules of registration login, personal home page, personal data, personal Weibo, personal photo album, attention to fans, group management and friend home page, the paper realizes the Sina Weibo friend recommendation system. The validity of the friend recommendation algorithm is verified by user data, and the goal of quickly helping users to filter effective information is achieved.
【学位授予单位】：燕山大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.3;TP393.092

【参考文献】