微博舆情分析系统信息处理模块的设计与实现
发布时间：2019-01-07 17:54
【摘要】：微博作为迅速崛起的新兴社交网络,因为信息的来源良莠不齐,以及民众的盲从性,利用微博传播谣言,扰乱社会治安的事件时有发生。如今,平均每小时的微博发布量高达数百万条,仅仅通过人工手段来对如此多条目进行监控和分析几乎是不可能完成的任务,因此依靠现代文本自动分析技术来开发一款微博舆情分析预警系统迫在眉睫。 本论文的工作是为一款微博舆情信息分析系统设计和开发其中的信息处理模块。论文首先介绍了微博舆情分析系统的整体框架设计,概括描述了该系统底层的信息采集、索引和分词模块和其所涉及的相关开源软件和技术。本系统的分析手段是通过微博关键词来进行的,对其所使用的潜在语义分析(LSA)也做出了相应的介绍。 论文的后几部分主要介绍信息处理模块,给出其整体设计架构,以及实时统计、自定义统计、同类词归并、微博影响力分析等功能点的具体设计和编码实现。系统完成的主要功能包括： 1)新浪微博的实时统计和预警； 2)各类复杂且精确地自定义统计 3)同类词归并； 4)微博分析以及用户分析； 5)提供分析功能API; 该系统在今后还将进一步发挥重要作用,为大政工平台的应用系统提供信息支持和数据共享。
[Abstract]:Weibo as a rapidly rising social network, because of the mixed sources of information, as well as the blindness of the public, the use of Weibo to spread rumors, disturbing social order incidents occur from time to time. Today, Weibo publishes millions of posts per hour on average. It is almost impossible to monitor and analyze so many items by manual means alone. Therefore, it is urgent to develop an early warning system for Weibo's public opinion analysis based on modern text automatic analysis technology. The work of this paper is to design and develop the information processing module of Weibo Public opinion Information Analysis system. Firstly, the paper introduces the whole frame design of Weibo public opinion analysis system, and describes the information collection, index and word segmentation module of the system and the related open source software and technology. The analysis method of this system is carried out by Weibo keyword, and the potential semantic analysis (LSA) used by Weibo is also introduced. The last part of the paper mainly introduces the information processing module, and gives its overall design framework, as well as the real-time statistics, custom statistics, congener word merging, Weibo influence analysis and other functional points of the specific design and coding implementation. The main functions of the system include: (1) real-time statistics and early warning of Sina Weibo; (2) complicated and accurate statistics (3) merging of similar words; (4) Weibo analysis and user analysis; 5) providing analysis function API; this system will further play an important role in the future, providing information support and data sharing for the application system of large political platform.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP311.52;TP393.092
本文编号：2403958
[Abstract]:Weibo as a rapidly rising social network, because of the mixed sources of information, as well as the blindness of the public, the use of Weibo to spread rumors, disturbing social order incidents occur from time to time. Today, Weibo publishes millions of posts per hour on average. It is almost impossible to monitor and analyze so many items by manual means alone. Therefore, it is urgent to develop an early warning system for Weibo's public opinion analysis based on modern text automatic analysis technology. The work of this paper is to design and develop the information processing module of Weibo Public opinion Information Analysis system. Firstly, the paper introduces the whole frame design of Weibo public opinion analysis system, and describes the information collection, index and word segmentation module of the system and the related open source software and technology. The analysis method of this system is carried out by Weibo keyword, and the potential semantic analysis (LSA) used by Weibo is also introduced. The last part of the paper mainly introduces the information processing module, and gives its overall design framework, as well as the real-time statistics, custom statistics, congener word merging, Weibo influence analysis and other functional points of the specific design and coding implementation. The main functions of the system include: (1) real-time statistics and early warning of Sina Weibo; (2) complicated and accurate statistics (3) merging of similar words; (4) Weibo analysis and user analysis; 5) providing analysis function API; this system will further play an important role in the future, providing information support and data sharing for the application system of large political platform.
【学位授予单位】：东华大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP311.52;TP393.092
【参考文献】
相关期刊论文 前9条
1 王娟;;网络舆情监控分析系统构建[J];长春理工大学学报(高教版);2007年04期
2 李林容;黎薇;;微博的文化特性及传播价值[J];当代传播;2011年01期
3 李红雷;巴明杰;;当前涉检网络舆情应对措施探析[J];中国检察官;2010年17期
4 朱颢东;钟勇;;结合优化的文档频和LSA的特征选择方法[J];计算机工程与应用;2009年34期
5 姚清耘;刘功申;李翔;;基于向量空间模型的文本聚类算法[J];计算机工程;2008年18期
6 刘克强;;2009共享版ICTCLAS的分析与使用[J];科教文汇(上旬刊);2009年08期
7 欧阳宏基;葛萌;赵蔷;;基于JDBC与设计模式的数据库连接池实现方法[J];计算机技术与发展;2011年01期
8 杨丽华;戴齐;郭艳军;;KNN文本分类算法研究[J];微计算机信息;2006年21期
9 程春蕊;刘万军;;高内聚低耦合软件架构的构建[J];计算机系统应用;2009年07期
，本文编号：2403958
本文链接：https://www.wllwen.com/guanlilunwen/ydhl/2403958.html
