基于HMM模型的网络社区推荐技术

摘要

由于网络社区相关数据具有海量、强噪音、实时变化性大等特点.因此,如何满足网络用户高质量和实时性的社区推荐需求,使得用户获得准确的推荐服务成为备受关注的研究热点.本文结合Web使用挖掘和内容挖掘的思想,提出了一种基于隐马尔可夫模型即HMM模型来描述用户访问社区模式的挖掘模型,并运用该模型来获取用户访问社区的具有共性的模式序列,经验证,该技术能够发现用户访问社区的迁移模式,并反映用户的访问偏好,从而将某社区内部成员共同感兴趣的新社区推荐给该社区的其他成员.%Because the data of the network community has the characteristics of large quantity,strong noise,real-time change and so on. Therefore,how to meet the requirements of high quality and real-time community recommendation of network users,and how users can get accurate recommendation service have become a hot research topic. In this paper,we propose a mining model based on Hidden Markov model(HMM),which is based on the idea of Web usage mining and content mining,and this model is used to obtain the common pattern sequence of user access community. It has been proved that this technology can discover the migration pattern of the user to visit the community,and reflect the user's access preference,so as to recommend a new community which is of common interest to the members of the community.