微博内容挖掘与金融时间序列关联性研究

发布时间：2018-05-27 03:42

本文选题：微博挖掘 + 情感分析　；参考：《北京邮电大学》2016年硕士论文

【摘要】：随着互联网时代的到来,社交媒体正日渐成为人们生活不可或缺的一部分。社交媒体为广大用户提供了一个即时的分享平台,用户可以通过文本、图片、音视频等方式在平台上分享消息,而平台的社交属性又使得这些消息可以沿着关注链呈指数级的传播。除了作为网民提供信息的重要平台,社交媒体也已经逐步成为网民获取信息的重要渠道。而伴随着移动互联网的兴起,微博凭借其面向移动用户的特点,得到了更加迅猛的发展。在全球范围内,80%的网民都在使用社交网络。中国互联网络信息中心报告显示,截止到2015年6月,中国网民达到了 6.68亿,其中,社交网络的使用率超过7成,而微博使用率则超过30%。2015年第三季度,微博的月活跃用户数已经超过2亿人,成为了中国最重要的社交媒体。微博的蓬勃发展带来了大量的内容信息,对于这些由用户生成内容进行挖掘,具有十分重要的意义。微博内容具备巨大的挖掘价值,这些信息一方面体现了用户对于自身生活状态、生活环境的态度,另一方面也包含了用户对于关乎国计民生的大事的意见和声音。通过对于微博的挖掘,提取出公众对于金融领域和事件的态度和情绪,厘清公众微博挖掘内容和金融问题的关系,对于个人和机构的投资决策,以及管理机构的政策制定都有重要的意义。对于微博平台的挖掘,一方面可以针对用户产生的内容进行挖掘,提取出用户在发布内容时传递出的潜在信息,另一方面,可以针对信息在传播过程中体现出来的用户关注关系网络进行挖掘。本论文基于上述思路,分别对微博进行了情感分析和图论分析,挖掘出微博文本和结构中包含的信息。为了将微博挖掘结果应用于金融领域,本论文设计了一种基于金融主题模型的关联算法,在微博与金融实体之间建立对应关系。依据对应关系,将微博挖掘结果与对应的金融时间序列进行关联性分析。根据关联性分析的结果,利用微博挖掘结果对金融时间序列进行分析和预测,并通过基于预测结果,实现自动交易策略,验证了微博内容挖掘结果对于金融时间序列预测的效果。
[Abstract]:With the advent of the Internet era, social media is increasingly becoming an indispensable part of people's lives. Social media provides a real-time platform for users to share messages on the platform through text, pictures, audio and video, etc. The social nature of the platform allows these messages to spread exponentially along the chain of concern. In addition to serving as an important platform for Internet users to provide information, social media has gradually become an important channel for Internet users to obtain information. With the rise of mobile Internet, Weibo has been developed more rapidly with its mobile-oriented characteristics. Around the world, 80% of Internet users are using social networks. According to the China Internet Network Information Center report, as of June 2015, the number of Chinese Internet users had reached 668 million, of which social network usage was more than 70 percent, while Weibo usage exceeded 30.2015 in the third quarter of 2015. Weibo has more than 200 million monthly active users, making it the most important social media in China. The vigorous development of Weibo brings a lot of content information, which is of great significance for the mining of user-generated content. The content of Weibo has great mining value. On the one hand, this information reflects the user's attitude towards their own living conditions and living environment, on the other hand, it also contains the users' opinions and voices on the major issues related to the national economy and the people's livelihood. Through the mining of Weibo, the attitudes and emotions of the public towards the financial field and events are extracted, the relationship between the contents of the public Weibo mining and financial problems is clarified, and the investment decisions of individuals and institutions are made. And the policy-making of the management organization has important significance. For the mining of Weibo platform, on the one hand, we can mine the content generated by the user, and extract the potential information that the user sends out when publishing the content, on the other hand, We can mine the user concern relation network which is reflected in the process of information dissemination. Based on the above ideas, this paper carries out affective analysis and graph theory analysis to Weibo, and excavates the information contained in Weibo text and structure. In order to apply the Weibo mining results to the financial field, this paper designs an association algorithm based on the financial subject model, and establishes the corresponding relationship between the Weibo and the financial entities. According to the corresponding relation, the correlation analysis between the Weibo mining results and the corresponding financial time series is carried out. According to the results of correlation analysis, the financial time series is analyzed and forecasted by using the results of Weibo mining, and the automatic trading strategy is realized based on the forecast results. The results of Weibo content mining are used to predict financial time series.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.1;TP393.092

【参考文献】