面向WEB的社会网络数据采集及分析平台

发布时间：2018-06-03 07:24

本文选题：社会网络 + 数据采集　；参考：《福州大学》2014年硕士论文

【摘要】：近年来,随着WEB技术和移动互联网技术的发展,以微博为代表的在线社交网站成为了人们日常交流、娱乐的重要工具。在线社交网站作为社交媒体,不仅具有新闻传播特性,也具有社交网络特性,可以从新闻话题传播、社会网络结构等多角度进行分析,具有很好的研究价值。同时,这些在线社交网站上用户的互动与交流,构建出结构复杂、数据规模庞大的社会网络,这给社会网络数据分析和研究带来了极大的挑战。为了更好地进行社会网络数据研究,本文针对基于WEB的在线社交网站,构建出社会网络数据采集及分析平台,并通过该平台,以微博社会网络为例,进行微博传播分析和微博用户关系分析。本文的研究内容主要有以下三个方面：首先,本文构建出社会网络数据采集及分析平台,提供在线社会网络数据的采集及分析功能,为大规模社会网络研究提供基础支持。平台整体采用分层的软件架构思想,保障系统的可扩展性。在数据采集方面,平台整合了不同的数据采集方式,同时针对不同数据类型分别进行采集和数据存储；在数据分析方面,平台对不同社会网络分析算法和可视化布局算法等进行封装,提供社会网络数据的可视化分析功能。接着,本文运用平台的数据采集和可视化分析功能,进行可视化微博传播分析。首先本文对微博社会网络中微博消息的传播特点进行分析,并利用平台采集不同数据样本进行对比实验,再运用可视化分析技术对样本中的微博消息传播网络进行分析,并获得微博在传播过程中的各项特征的统计数据,发现关键传播帐号在微博传播过程中起到的巨大作用。最后,基于平台对大规模社会网络数据的处理能力,本文对影响微博社会网络用户关系的相关特征进行分析。首先本文分析了网络结构特征在微博社会网络中的作用,同时引入微博属性特征,构造出基于随机森林的链路预测模型。运用该模型对平台所采集的数据集进行训练预测,通过比较引入微博属性特征前后的预测性能以及特征的重要性分布,分析了各类特征对微博用户关系形成的影响,揭示了除传统的网络结构特征外,微博属性特征对微博用户关系的形成具有重要的影响力。
[Abstract]:In recent years, with the development of WEB technology and mobile Internet technology, the online social network represented by Weibo has become an important tool for people's daily communication and entertainment. As social media, online social network not only has the characteristics of news dissemination, but also has the characteristics of social network. It can be analyzed from news topic communication, social network structure and so on, which has good research value. At the same time, the interaction and exchange of users on these online social networking sites have created a complex social network with large data scale, which brings great challenges to the analysis and research of social network data. In order to study social network data better, this paper constructs a social network data collection and analysis platform for online social network based on WEB, and takes Weibo social network as an example. Weibo propagation analysis and Weibo user relationship analysis. The main contents of this paper are as follows: firstly, this paper constructs a social network data acquisition and analysis platform to provide online social network data collection and analysis function, providing basic support for large-scale social network research. The whole platform adopts the idea of layered software architecture to ensure the expansibility of the system. In the aspect of data acquisition, the platform integrates different ways of data acquisition, at the same time, it collects and stores data according to different data types, and in the aspect of data analysis, The platform encapsulates different social network analysis algorithms and visual layout algorithms to provide visual analysis function of social network data. Then, this paper uses the function of data acquisition and visual analysis of the platform to analyze the visual Weibo propagation. Firstly, this paper analyzes the characteristics of Weibo message transmission in Weibo social network, and uses the platform to collect different data samples to carry on the contrast experiment, and then uses the visual analysis technology to analyze the Weibo message dissemination network in the sample. The statistical data of the characteristics of Weibo in the process of communication are obtained, and the key communication account plays an important role in the process of Weibo propagation. Finally, based on the platform's ability to deal with large-scale social network data, this paper analyzes the related features that affect the user relationship of Weibo social network. Firstly, the function of network structure in Weibo social network is analyzed, and the Weibo attribute feature is introduced to construct a link prediction model based on stochastic forest. The model is used to train and predict the data set collected by the platform. By comparing the prediction performance before and after the introduction of the Weibo attribute feature and the importance distribution of the feature, the influence of various features on the formation of Weibo user relationship is analyzed. In addition to the traditional network structure features, the Weibo attribute features play an important role in the formation of Weibo user relationships.
【学位授予单位】：福州大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP274.2;TP393.09

【相似文献】