社会化标签系统中基于多维社会网络的Web知识推送研究

发布时间：2018-05-02 00:01

本文选题：社会化标签 + 多维社会网络　；参考：《华中师范大学》2017年硕士论文

【摘要】：在Web2.0时代,互联网在成为世界上最大的知识库以给人们带来巨大便利的同时,也引发了 “知识迷航”、“知识泛滥”等问题。近年来兴起的Web知识推送能够解决这些问题,其以Web用户在互联网上的行为记录为基础,挖掘Web用户的各种知识需求,然后再整合相关资源或者知识主动提供给用户,以满足个性化需求。本文主要以多维社会网络分析为切入点尝试对这一领域中的模型进行探索,以优化Web知识推送的效果。本文首先从内涵、特点、模型以及功能等方面对社会化标签系统、社会网络分析与Web知识推送相关理论进行了阐述。然后,提出多维社会网络的构建方法,科学“移植”借鉴单维社会网络分析原理探索形成个体级、局部级和全局级三个层次的多维社会网络分析框架。其次,探索基于标签主题的多维社会网络的内涵,并使用多维社会网络分析的分析指标逐层分析。再次,基于多维社会网络构建Web知识推送模型,利用多维社区发现结果与协同过滤思想计算Web用户对知识资源的兴趣度,并最终产生Web知识推送集。最后,结合真实数据集,构建并基于标签主题的多维社会网络,依据Web知识推送模型完成Web知识推送,经过对Web知识推送结果的精确度评价,证明了本文提出方法的可行性和有效性。
[Abstract]:In the era of Web2.0, the Internet has become the largest knowledge base in the world to bring people great convenience, but also caused "knowledge confusion", "knowledge overflow" and other problems. In recent years, Web knowledge push can solve these problems. Based on the behavior records of Web users on the Internet, it excavates various knowledge needs of Web users, and then integrates related resources or knowledge to provide users actively. To meet the individual needs. In order to optimize the effect of Web knowledge push, this paper attempts to explore the model in this field by using multidimensional social network analysis as the starting point. In this paper, the theory of social label system, social network analysis and Web knowledge push is first expounded from the aspects of connotation, characteristics, model and function. Then, the construction method of multi-dimensional social network is put forward, and the scientific "transplantation" is used for reference to explore the analysis framework of multi-dimensional social network at individual, local and global levels with reference to the analysis principle of one-dimensional social network. Secondly, the connotation of multidimensional social network based on label theme is explored, and the analysis index of multi-dimensional social network is analyzed layer by layer. Thirdly, the Web knowledge push model is constructed based on multidimensional social network, and the interest of Web users to knowledge resources is calculated by using multi-dimensional community discovery results and collaborative filtering ideas, and finally the Web knowledge push set is generated. Finally, a multi-dimensional social network based on label topic is constructed and based on the real data set. According to the Web knowledge push model, the Web knowledge push is completed, and the accuracy of the Web knowledge push result is evaluated. The feasibility and effectiveness of the proposed method are proved.
【学位授予单位】：华中师范大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.09;G354

【参考文献】