社交网络海量数据的分析与可视化
发布时间：2019-02-25 18:22
【摘要】：在社交网络中,每个用户都是一个数据源,其数据量成指数级爆炸性增长,而其中蕴含的价值是不言而喻的。在如今的信息社会中,能否更多的掌握信息是抢占市场的关键。对社交网络的海量数据进行分析与可视化,正是对这些数据提取和运用的重要过程。 数据可视化技术一直以来是学术界研究的热点,它可以将抽象的数据变成人们便于理解和观察的图形,能够直观的表达出数据中的信息和意义。近年来,当数据可视化遇到大数据时,是数据可视化领域的又一次挑战。本文针对这一问题提出了一种基于社团发现的多层级可视化布局算法,通过测试和对比,验证其有效的降低了计算复杂度,增强了对海量数据展现的容纳能力。 基于上述提出的算法,本文还实现了一个社交网络信息可视化系统。该系统集成了数据爬取模块,数据格式化模块,社团发现模块,文本信息向量化模块以及可视化模块,可以展现出海量社交网络的网络结构以及信息分布,并具有较强的数据承载能力。本文首先介绍了系统的需求分析与总体设计,然后介绍了系统各个模块的具体实现,最后对系统做了完整的测评和分析,并提出下一步的工作展望。
[Abstract]:In social networks, each user is a data source, and the amount of data increases exponentially, and its value is self-evident. In today's information society, whether to grasp more information is the key to preemptive market. Analyzing and visualizing the massive data of social network is the important process of extracting and applying these data. The technology of data visualization has always been a hot topic in academia. It can transform abstract data into graphics that people can easily understand and observe, and can express the information and meaning of data intuitively. In recent years, when data visualization meets big data, it is another challenge in the field of data visualization. In this paper, a multi-level visual layout algorithm based on community discovery is proposed. Through testing and comparison, it is proved that the algorithm effectively reduces the computational complexity and enhances the capacity of mass data display. Based on the proposed algorithm, this paper also implements a social network information visualization system. The system integrates data crawling module, data formatting module, community discovery module, text information vectorization module and visualization module, which can show the network structure and information distribution of mass social network. And has a strong data bearing capacity. This paper first introduces the requirements analysis and overall design of the system, then introduces the concrete implementation of each module of the system. Finally, it makes a complete evaluation and analysis of the system, and puts forward the future work prospects.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.41;TP393.092
本文编号：2430414
[Abstract]:In social networks, each user is a data source, and the amount of data increases exponentially, and its value is self-evident. In today's information society, whether to grasp more information is the key to preemptive market. Analyzing and visualizing the massive data of social network is the important process of extracting and applying these data. The technology of data visualization has always been a hot topic in academia. It can transform abstract data into graphics that people can easily understand and observe, and can express the information and meaning of data intuitively. In recent years, when data visualization meets big data, it is another challenge in the field of data visualization. In this paper, a multi-level visual layout algorithm based on community discovery is proposed. Through testing and comparison, it is proved that the algorithm effectively reduces the computational complexity and enhances the capacity of mass data display. Based on the proposed algorithm, this paper also implements a social network information visualization system. The system integrates data crawling module, data formatting module, community discovery module, text information vectorization module and visualization module, which can show the network structure and information distribution of mass social network. And has a strong data bearing capacity. This paper first introduces the requirements analysis and overall design of the system, then introduces the concrete implementation of each module of the system. Finally, it makes a complete evaluation and analysis of the system, and puts forward the future work prospects.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP391.41;TP393.092
【参考文献】
相关期刊论文 前7条
1 樊鹏翼;王晖;姜志宏;李沛;;微博网络测量研究[J];计算机研究与发展;2012年04期
2 李克潮;梁正友;;适应用户兴趣变化的指数遗忘协同过滤算法[J];计算机工程与应用;2011年13期
3 王柏;吴巍;徐超群;吴斌;;复杂网络可视化研究综述[J];计算机科学;2007年04期
4 孙扬;蒋远翔;赵翔;肖卫东;;网络可视化研究综述[J];计算机科学;2010年02期
5 台德艺;王俊;;文本分类特征权重改进算法[J];计算机工程;2010年09期
6 吴鹏;李思昆;;适于社会网络结构分析与可视化的布局算法[J];软件学报;2011年10期
7 肖有诰;屠成宇;;基于启发式函数的分布式FN算法[J];计算机系统应用;2012年10期
相关博士学位论文 前1条
1 孙扬;多变元网络数据可视化方法研究[D];国防科学技术大学;2010年
，本文编号：2430414
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2430414.html
