摘要随着互联网的日益发展,社交类网站已经渗透我们日常生活的很多角落,随之而产生了巨大的信息量。而寻找到社交网络的核心用户,有助于我们对这些社交网络进一步分析。我将从两个角度进行探索:1.从比较粉丝数等数据的层面进行分析。2.从通过评估用户影响力的抽象层面进行分析。在数据层面,利用邻接表存储结构来构成用户间的关系图,利用遍历搜索和排序算法来分析各项数据。在抽象层面,从用户通过社交网络和其他用户互动的过程中,把抽象的,深层的信息数据化,图像化,发现局部社区,计算用户相似性程度。例如采用最短路径法,对某些用户的共同关注进行向上搜索。本研究采用了很多基础算法以及数据挖掘中的算法,建立了一个对寻找社交网络核心用户的概要设计。65599
毕业论文关键字: 数据挖掘 遍历搜索 最短路径法 相似性程度
毕业设计说明书(论文)外文摘要
Title Social Networking Core Community Members Mining Algorithm
Abstract
With the increasing development of the Internet, social networking sites have penetrated many corners of our daily lives, the attendant produced a huge amount of information. And looking into the core of social network users, help us to further analyze these social networks. I will explore two perspectives: one from the number of fans and other data comparing the level of analysis. 2 From the user by evaluating the influence of level of abstraction for analysis. At the data level, the use of the adjacent table storage structure to form the relationship between users diagram using traversal search and sort algorithms to analyze the data. In the abstract level, from a user through social networks and other user-interactive process, the abstract, deep information data, image technology and found that the local communities, the degree of similarity computing users. Such as using the shortest path method, a common concern for some users for up searching. This study uses a lot of basic algorithms and data mining algorithms, the establishment of a social network for finding the core user's profile designs.
目录
1 绪论 2
1.1 选题的背景和意义 2
1.2 研究的目的和主要任务 4
2 社交网络和复杂网络 5
2.1 关于度和度分布的研究 5
2.2 关于中心性的分析 5
2.3 关于核心-缘结构的分析 6
2.4 网络的重叠群落结构 6
3 发现社区和寻找核心用户 7
3.1 局部网络发现算法 7
3.2 通过比较粉丝数来寻找核心用户。 9
3.3 最短路径算法 11
3.4 用户的相似性程度和用户影响力 13
4 对两个层次的比较 17
4.1 粉丝的价值