面向用户体验的智能应用使用模式与优化的研究

发布时间：2019-04-13 07:52

【摘要】：当今时代,移动互联网正在快速的发展,与此同时,大数据时代也悄然而至,“移动互联网”和“大数据”变成当今当前互联网领域内最火的话题。而在这之中,和用户关系最直接的移动互联网便是移动APP,而面对海量的APP,如何选择合适的APP,对普通用户来说是一个头疼的问题。在这种情况下,如何在大量数据里帮助用户选择合适的APP,以提升用户的使用体验并且能为用户节约使用成本变得重要起来。基于这个情况,本文将从流量使用方面,结合用户的使用偏好,为用户推荐符合自身使用习惯并且减少流量使用的APP。本文首先研究了如何进行数据分析平台的搭建,实现了基于Ambari的Hadoop分析平台;在此基础上对目标数据集进行相关的分类、处理,进而对数据做一些相关的分析。其次,本文基于APP的流量消耗和流行度建立APP推荐模型,并结合用户相关的使用偏好研究APP推荐模型。根据建立好的推荐模型,分析出用户的使用偏好,为用户推荐相似的但流量更少,流行度更高的APP,改善用户的使用体验。除此之外,不论是用户还是APP,都需要考虑时间段的问题,也就是说用户偏好使用某一类APP的时间段和APP被使用最频繁的时间段。最后,本文依据移动互联网用户相关的数据集,为上述APP推荐模型进行了验证,结果表明在满足用户的使用偏好情况下,为用户推荐的APP能够比用户原来使用的APP花费更少的流量或者推荐的APP具有更高的流行度。这样一来达到了提高用户使用体验的目的。
[Abstract]:Nowadays, the mobile Internet is developing rapidly. At the same time, the era of big data has come quietly. "Mobile Internet" and "big data" have become the hottest topics in the current Internet field. Among these, the most direct relationship with the mobile Internet is the mobile APP, and facing the mass of APP, how to choose the right APP, is a headache for ordinary users. In this case, it is important to help users choose the right APP, in large amounts of data to improve the user's experience and save users the cost of use. Based on this situation, this article will from the traffic use aspect, combines the user's use preference, recommends to the user the APP. which conforms to its own usage habit and reduces the traffic use. Firstly, this paper studies how to build the data analysis platform, and realizes the Hadoop analysis platform based on Ambari. On this basis, it classifies and processes the target data set, and then makes some correlative analysis to the data. Secondly, based on the traffic consumption and popularity of APP, this paper establishes the APP recommendation model, and studies the APP recommendation model based on user-related usage preferences. According to the established recommendation model, the users' preference is analyzed, and the similar APP, with less traffic and higher popularity is recommended for users to improve the user's experience. In addition, both users and APP, need to consider the problem of time periods, that is, users prefer to use a certain kind of APP period and APP is used most frequently. Finally, according to the data sets related to the mobile Internet users, this paper validates the APP recommendation model mentioned above. The results show that in the case of satisfying the user's usage preference, The APP recommended for the user can cost less traffic or the recommended APP has a higher popularity than the original APP used by the user. In this way, the user experience has been improved.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP311.56

【参考文献】