手机游戏推荐技术研究
时间:2018-09-10 14:31 来源:毕业论文 作者:毕业论文 点击:次
| 摘要伴随着计算机技术和互联网技术的发展，互联网为用户提供了越来越便捷的服务。在21世纪里，智能手机的出现，手机信息也在进入一个爆炸的时代。手机的快速发展，手机游戏变得越来越丰富和火热，手机游戏中海量信息意着巨大的利润。收集手机用户的历史行为和数据，建立起相关学习模型，挖掘用户的兴趣和需求，并从数据模型中，分析并预测用户的购买集，搭建一个推荐系统，是每个手机游戏的开发商与发行商所希望做的。28079 在本课题中，所采用的数据是由某游戏公司所提供的手机游戏用户数据集。根据数据集中所提供的用户信息，用户频繁登录信息以及用户支付信息，分析并筛选出其有用的用户信息。预测其用户可能的兴趣行为，并根据用户的可能行为，对用户推荐其接受并可能支付的手机计费游戏。在目前的推荐算法中，有基于规则、基于内容等主流的推荐算法。其中，协同过滤是这些算法中最为经典且有效的推荐算法。本课题采用协同过滤算法来设计推荐系统，并针对其表现出的问题如数据稀疏、扩展性差等问题，做一些对应的改进，并结合多种最近邻算法融合模型来获取最优推荐。 毕业论文关键词 推荐系统 协同过滤 数据稀疏 外文摘要 Title The Research on the Recommendation Technology for the Mobile Games Abstract With the development of computer technology and network technology, Internet has bring more and more convenience service for user. In the 21st century, the emergence of the smart phone, mobiles phone information has entered into a exploding period. With the rapid development of mobiles phone, mobile games become more and more various and hot, Huge data information In mobile games also means big profit. Collecting historical behavior and behavior of mobile phone user, establish a learning model, and mining user possible user interest and need, analyze and predict user buy set, and establish this recommendation system, is every mobile In our project, out dataset is the mobile game user dataset provided by Duohe Game company. Based on the user information, user login record and user game order record in dataset, analyse and filter useful user information. Predict user possible behavior and mobile charging game which user will accept and paid to user on the basis of the user possible behavior. Nowadays, there are many recommend main stream algorithm like based on rule, based on content and others. One of them is every mobile games developer and publisher hope to do. collaboration filtering which is also the most classical and the most effective recommendation algorithm. Our project aim to use collaboration filter to establish the recommendation system. We will contrary to some possible problems like data sparsity and bad expansibility and do some corresponding solutions, and combine model by some nearest neighbor methods to get best answer. Keywords recommendation system, collaborative filtering, data sparsity 目 录 中文摘要 3 外文摘要 4 目 录 I 1 绪论 1 1.1 研究背景与意义 1 1.2 国内外研究现状 1 1.3 本文主要内容及结构安排 1 2 协同过滤推荐技术概述 3 2.1 常见推荐技术 3 (责任编辑：qin) |