基于聚类算法的吉林大学校园新闻推荐系统的设计与实现

发布时间：2018-06-08 18:53

本文选题：推荐系统 + 聚类算法　；参考：《吉林大学》2017年硕士论文

【摘要】：近几年来的时代最热点无疑是移动互联网时代的到来,互联网化的浪潮席卷了现代社会的各个领域。随着移动通讯工具的普及以及互联网信息技术的发展,“信息过载”的现象已经成为无法避免的焦点问题。“信息过载”的挑战是人人无法回避的,无论是对于信息的发布者,还是对于信息的接收者,都必须去面对这个问题。对于信息的接受者尤甚,如何快速地获得自身需要的,有价值的信息值得关注。由于以上原因,个性信息推荐成为了计算机领域的一个研究热点。校园信息化是社会信息化的一个重要领域,信息过载和信息获取效率的问题同样突出。当前校园内学生获取校园内新闻通知的方式极为多样化,却缺乏效率,缺少精准性和智能性。本文围绕以上背景,关注校园信息化建设,以及怎样做一个校园新闻推荐系统做详细的分析与研究。本文的研究内容有如下:文章通过对校园推荐系统的研究现状与发展进行分析,以此提出了未来的个性化推荐发展方向将是结合各个推荐方法优点的组合推荐方法。对现有的个性化推荐方法如基于关联规则推荐,基于内容推荐,协同过滤推荐等进行了研究,分析其中的技术原理,应用领域和各自优缺点。结合对现有推荐方法的研究分析,本文提出了一种先基于文本聚类,再将内容推荐和协同过滤结合的推荐方法。因此本文采用K-means算法对新闻利用关键词特征进行新闻文本聚类处理;在关注文本内容相似性的同时,结合对用户行为的协同过滤分析进行推荐。基于本文提出的推荐方案,设计并实现了一个基于Android平台的校园信息推荐系统,应用于聚类及推荐技术,新闻采集的实现,服务器的搭建,数据库及Android客户端的实现并给出了测试与验证。
[Abstract]:In recent years, the most hot spot of the times is undoubtedly the arrival of the mobile Internet era, the wave of Internet swept through all fields of modern society. With the popularity of mobile communication tools and the development of Internet information technology, the phenomenon of "information overload" has become an unavoidable focus. The challenge of "information overload" is unavoidable, both for the publisher and receiver of information. Especially for the recipients of information, how to quickly obtain their own needs, valuable information is worth paying attention to. Because of the above reasons, personality information recommendation has become a research hotspot in the field of computer. Campus informatization is an important field of social informatization. The problems of information overload and information acquisition efficiency are also prominent. At present, students in campus get news notification in a variety of ways, but lack of efficiency, accuracy and intelligence. This paper focuses on the construction of campus information, and how to do a campus news recommendation system for detailed analysis and research. The research contents of this paper are as follows: through the analysis of the current situation and development of campus recommendation system, this paper proposes that the future development direction of personalized recommendation will be combined with the advantages of each recommendation method. The existing personalized recommendation methods, such as association rule based recommendation, content based recommendation and collaborative filtering recommendation, are studied, and their technical principles, application fields and their respective advantages and disadvantages are analyzed. Based on the research and analysis of the existing recommendation methods, this paper proposes a recommendation method based on text clustering, and then combines content recommendation with collaborative filtering. In this paper, K-means algorithm is used to deal with the news text clustering using keyword features, while paying attention to the similarity of text content, combining with the collaborative filtering analysis of user behavior to recommend. Based on the recommendation scheme proposed in this paper, a campus information recommendation system based on Android platform is designed and implemented, which is used in clustering and recommendation technology, news collection and server construction. The implementation of database and Android client is given.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.3

【参考文献】