基于情境感知的移动终端用户消费行为预测研究

发布时间：2018-08-27 09:50

【摘要】：随着移动互联网和移动计算技术不断发展,智能手机等移动终端成为人们通信交流和数据访问的主要工具,情境感知服务在这样的背景下产生并日渐影响着人们的生活。与此同时,人们生活水平的提高,消费购物已经成为人们生活的重要组成部分。根据用户当前的情境,对用户即将进行的行为和需要的服务进行预测,在企业的精确化营销、商品服务的个性化定制方面有较大的意义,同时也能为用户节约时间和精力,营造良好的便捷消费体验。目前,在基于情境的消费行为预测方面的研究还处于初步阶段,存在情境权重缺乏量化标准,用户行为预测准确性不足等问题。如何有效获取情境信息,并基于情境对用户的消费行为进行预测成为了个性化服务领域面临的课题。本课题依托国家自然科学基金项目,通过研究移动环境下用户情境信息的获取和分析策略,以及基于情境的用户消费行为预测方法,从而挖掘有效信息,使用户能够通过移动终端更好地体验并获取情境感知所带来的个性化推荐等优质服务。本文主要研究工作如下:①分析国内外在情境感知和移动终端行为分析领域的研究现状,并针对目前存在的问题提出解决方案。②探究情境感知的理论和技术,分直接情境信息采集和潜在情境信息获取两方面研究移动终端用户情境信息获取的方法,并对获取的情境信息进行量化模型构建。③通过研究情境建模方法以及情境相似度算法,对情境信息进行建模、分析和处理。在此基础上结合情境信息熵的思想对情境权重进行计算,以便提高情境相似度计算的准确性。④结合情境感知技术和移动终端用户消费行为的特殊性,改进传统的用户行为分析和预测算法,提出一种基于情境相似度和二次聚类的协同过滤算法(CTCF)和一种基于情境的关联规则算法,并进行仿真实验验证其效果和性能。根据这两种算法的侧重点,提出基于这两种算法的组合预测机制,增加用户行为预测的准确性和实用性。⑤结合商业领域移动终端用户消费行为预测的解决方案,设计一个基于用户消费行为预测的自助导购系统,并将本文的预测算法应用于该系统的消费推荐中。
[Abstract]:With the development of mobile Internet and mobile computing technology, mobile terminals such as smart phones have become the main tools of communication and data access. At the same time, with the improvement of people's living standard, consumer shopping has become an important part of people's life. According to the user's current situation, the prediction of the user's upcoming behavior and needs will have great significance in the enterprise's precise marketing and the personalized customization of the goods and services, and at the same time, it can save time and energy for the user. Create a good experience of convenient consumption. At present, the research on situation-based consumer behavior prediction is still in the initial stage, there are some problems such as the lack of quantitative standard of situation weight, the lack of accuracy of user behavior prediction and so on. How to obtain the situation information effectively and predict the consumer behavior based on the situation has become a topic in the field of personalized service. Based on the project of National Natural Science Foundation of China, this paper studies the strategies of obtaining and analyzing user's situation information in mobile environment, as well as the forecasting method of user's consumption behavior based on situation, so as to mine the effective information. Through mobile terminals, users can better experience and obtain the high quality services such as personalized recommendation brought about by situational awareness. The main research work of this paper is as follows: 1. Analyze the current situation of situational perception and mobile terminal behavior analysis at home and abroad, and propose a solution for the existing problems. 2. Explore the theory and technology of situational awareness. The methods of situation information acquisition for mobile terminal users are studied in two aspects: direct situation information acquisition and potential situation information acquisition. By studying the situational modeling method and the situational similarity algorithm, the situational information is modeled, analyzed and processed. In order to improve the accuracy of situational similarity calculation, combining with context-aware technology and the particularity of consumer behavior of mobile end-users, the paper calculates the weight of situation with the idea of situational information entropy. By improving the traditional user behavior analysis and prediction algorithm, a collaborative filtering algorithm based on situational similarity and quadratic clustering (CTCF) and a situation-based association rule algorithm (CTCF) are proposed, and simulation experiments are carried out to verify the effectiveness and performance of the algorithm. According to the emphasis of the two algorithms, the combined prediction mechanism based on these two algorithms is proposed to increase the accuracy and practicability of user behavior prediction, and combine with the solution of consumer behavior prediction of mobile end users in the commercial field. A self-help shopping guide system based on consumer behavior prediction is designed, and the prediction algorithm is applied to the consumption recommendation of the system.
【学位授予单位】：重庆大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP391.3

【相似文献】