网络行为特征模型及在个性化服务中的应用

发布时间：2018-04-10 07:55

本文选题：网络行为特征　切入点：兴趣感知　出处：《华南理工大学》2013年硕士论文

【摘要】：随着互联网技术的发展，互联网资源成为人们获取资源的重要手段。为了降低搜索成本，先后出现了分类网站、搜索引擎等提高资源搜索效率的技术，这些技术在一定程度上缓解了互联网海量信息与用户获取互联网信息手段单一、简单的矛盾，但由于互联网资源以指数级速度增长，用户从海量信息中获取自己所需的仍需花费较大的代价。基于用户兴趣感知的互联网资源个性化推送技术和搜索技术是互联网资源有效获取的新方法，而如何通过对用户行为的分析挖掘用户兴趣，则是实现互联网资源个性化推送和搜索的关键。本文提出了一种基于网络行为特征模型的Web个性化服务系统，该系统以网络行为特征模型为基础，根据用户兴趣有针对性的推送给用户不同的网页。本文首先对用户行为进行收集、分析，对现有一些网络行为模型进行研究，针对它们的不足，提出首先对用户网络行为特征进行分类，然后根据不同类型的用户行为建立其网络行为特征模型，，以此预测用户单页面兴趣度，并对模型进行检验，说明本文所选模型的合理性。文章将该网络行为特征模型应用到个性化服务当中，根据网络行为特征模型得到用户感兴趣的文档集，通过对该文档集进行挖掘形成树状结构的用户兴趣集。最后本文在Android平台上设计了一个基于网络行为特征模型的个性化服务应用系统，该系统结合用户兴趣集与用户地理位置，提供个性化服务。本论文提出的个性化服务系统可为以后Web个性化服务提供一定的参考，并且可以应用于如电子商务、个性化浏览器等其他领域。
[Abstract]:With the development of Internet technology, Internet resources become an important means for people to obtain resources.In order to reduce the search cost, several technologies such as classified websites and search engines have emerged to improve the efficiency of resource search. To some extent, these technologies have alleviated the contradiction between the huge amount of information on the Internet and the simple and simple means for users to obtain Internet information.However, due to the exponential growth of Internet resources, it is still costly for users to obtain what they need from mass information.Personalized push and search technology of Internet resources based on user interest perception is a new method to obtain Internet resources effectively.It is the key to realize personalized push and search of Internet resources.In this paper, a Web personalized service system based on network behavior feature model is proposed. The system is based on the network behavior feature model and can be pushed to different web pages according to users' interests.In this paper, the user behavior is collected, analyzed, and some existing network behavior models are studied. In view of their shortcomings, this paper proposes a classification of user network behavior characteristics.Then, according to different types of user behavior, the network behavior characteristic model is established to predict the single page interest of users, and the model is tested to illustrate the rationality of the model selected in this paper.In this paper, the network behavior feature model is applied to the personalized service. According to the network behavior feature model, the document set of user interest is obtained, and the user interest set with tree structure is formed by mining the document set.Finally, this paper designs a personalized service application system based on network behavior feature model on Android platform, which combines user interest set and user geographical location to provide personalized service.The personalized service system proposed in this paper can provide some reference for the future Web personalized service, and can be applied to other fields such as e-commerce, personalized browser and so on.
【学位授予单位】：华南理工大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.3

【参考文献】