面向移动终端用户的推荐系统研究

发布时间：2018-01-07 22:21

本文关键词：面向移动终端用户的推荐系统研究　出处：《河南大学》2015年硕士论文　论文类型：学位论文

【摘要】：近年来,随着互联网的快速传播,人们越来越多的商业行为也从实际生活转移到了网络生活上。现在电子商务在我们的生活中飞快的发展,更多的产品在互联网上供人们选择,普通的筛选方式(搜索引擎)已经难以满足人们的需求了,这就促进了推荐系统的产生。推荐系统是系统通过计算用户的历史信息,给用户推荐适合的产品。随着智能手机用户的快速增加,人们进入到了一个应用电脑办公,应用智能手机娱乐的时代。智能手机用户的增多促进了智能化应用软件的发展,也使得越来越多的电子商务和娱乐平台将重心转移到了智能手机用户中来,所以基于移动终端用户的推荐系统需要得到进一步的研究。现今时代是一个大数据的时代,如何从海量的呈爆炸式增长的数据中,快速有效的找到自己感兴趣的东西,是基于智能终端用户的推荐系统需要面临的问题。本文以传统推荐系统的研究现状为出发点,通过对传统推荐系统的研究和对智能终端的特点研究,针对智能终端的特殊性,根据智能终端用户的需求改善了传统的推荐算法,在研究过程中对协同过滤算法、多样化算法进行了改进。本文的主要工作体现在一下几个方面:(1)提高用户冷启动的准确性:针对移动用户的移动化特点,对传统的协同过滤算法进行了改进,将移动用户的情景变化考虑进去,计算项目的相似度并进行推荐。(2)提高推荐多样性:将推荐数目进行划分,将移动用户的情景考虑在内,提高推荐多样的多样性。(3)提高系统的可扩展行:应用MapReduce,并使用Hadoop架构以改善大量用户和产品的加入问题,提升了推荐系统的可扩展性。
[Abstract]:In recent years, with the rapid spread of the Internet, business more and more people from the real life to network life. Now fast development of electronic commerce in our life, more products on the Internet for people to choose, the ordinary way of screening (search engine) has been difficult to meet the needs of the people, the on the promotion of recommendation system. Recommender system by calculating the user history information, recommend suitable products. With the rapid increase of intelligent mobile phone users, people enter into a computer office application, application of intelligent mobile phone entertainment era. More intelligent mobile phone users to promote the development of intelligent application software also, the electronic commerce and entertainment platform more and more shifted to the intelligent mobile phone users, so the recommendation system based on the need of the user's mobile terminal Further research. This era is an era of big data, how from the mass of the explosive growth of data, quickly and efficiently find what they are interested in, is the need to face the intelligent recommendation system based on the problem of end users. Based on the traditional recommendation system research system as a starting point, through the study on the traditional recommendation system and characteristics of research on intelligent terminal, aiming at the particularity of the intelligent terminal, according to the needs of users of intelligent terminal to improve the traditional recommendation algorithm, in the process of research on collaborative filtering algorithm, various algorithms are proposed. The main work is reflected in the following aspects: (1) improve the accuracy user cold start: for the mobile characteristic of mobile users, collaborative filtering algorithm to improve traditional scene change the mobile user into account, calculation of similar projects Recommended and. (2) improve the recommendation diversity: the recommended number of mobile users will be divided, the situation into account, improve the diversity of recommended variety. (3) improve the scalability of the systems: application of MapReduce, and use the Hadoop framework to improve the large number of users and products to enhance the problem. The scalability of the recommendation system.

【学位授予单位】：河南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.3

【相似文献】