项目名称: 通过社会化媒体挖掘用户兴趣的方法及应用研究
项目编号: No.71272029
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 管理科学
项目作者: 刘红岩
作者单位: 清华大学
项目金额: 60万元
中文摘要: 近年来,社交网络、信息分享平台等社会化媒体的迅猛发展不仅改变了人与人之间的交流方式,也改变了企业的经营模式,使企业的经营更依赖于对大量数据的分析与发现。本研究旨在通过微博这类社会化媒体挖掘用户的行为模式,多方位描述用户兴趣,从而挖掘社会化媒体中的巨大商业价值。已有的用户兴趣的表达和挖掘方法不足以处理微博平台中海量的短文本、动态变化的大规模网络以及富媒体、非结构化的数据,也不能充分挖掘其中隐含的新颖行为模式。本项目的研究内容包括:用户在社会化媒体中行为模式的挖掘方法研究、用户兴趣建模研究、用户兴趣推测方法研究、典型用户群识别及差异研究、基于用户兴趣的典型应用研究。该研究将在话题识别、信息传播、意见挖掘以及行为预测等方面解决关键的科学问题,提出一套具有创新性的知识发现理论、模型和方法,并开发基于用户兴趣的应用演示系统,为电子商务中的个性化推荐、开发新型商业模式及社会化营销提供理论和技术支持。
中文关键词: 社会化媒体;用户兴趣;数据挖掘;社会网络;文本挖掘
英文摘要: In recent years, social media such as social network and information sharing platform are undergoing rapid development, which has not only changes our ways of information exchange, but also changes the ways of doing business and make the business more depend on analysis and discovery from large volume of data. In this project, we want to study user's behavior patterns through social media in order to describe user's interests from multiple perspectives, and mine the huge business value behind social media. Existing methods proposed to describe user's interests are not enough to deal with huge amount of short texts, dynamic changing network,multimedia and non-structured data, and cannot discover all of new user behaviors hidden in social media. Major research content includes research on the mining of user behavior patterns, research on the user interest modeling, research on the inferring of user interest, research on the detection of typical user groups and difference study, and research on the typical application based on user interests. We will solve important scientific problems about topic detection, information propagation, opinion mining and behavior prediction, and propose a set of theorems, models and methods in the knowledge discovery area. We will also develop demonstration application system based on
英文关键词: social media;user interest;data mining;social network;text mining