In the era of rapid development of the Internet, in order to reflect the evolution process of users' viewpoints on network relations, a Bayesian viewpoint evolution model based on Weibo data mining is proposed by studying the relationship between the viewpoints of the author and those of the forwarders on the Sina Weibo platform. Firstly, Python crawler technology was used to crawl the comments and forwarding data under the Weibo topic “#ChatGPT father says human-level AI is coming soon”. After data preprocessing and sentiment analysis, the user relationship network diagram was drawn with Gephi software. Secondly, the viewpoint evolution model is constructed and the viewpoint update formula based on Bayes rule is used to calculate the users' viewpoint evolution within the network relations of several kinds of different publication centers. The results show that: in the communication of public opinion, the evolution direction of the opinions of the media-centered network relations tends to be more consistent, which indicates the importance of the opinion guidance of the media in the communication of information. The analysis and technology provide a certain reference for the government and the media to control and guide the network public opinion.

中文翻译：

---

基于微博数据挖掘的视点演化分析


在互联网快速发展的时代，为了反映用户视点对网络关系的演化过程，通过研究笔者与新浪微博平台上转发者视点的关系，提出了一种基于微博数据挖掘的贝叶斯视点演化模型。首先，使用 Python 爬虫技术对微博主题“#ChatGPT 爸爸说人类级 AI 即将推出”下的评论进行爬取和转发数据。经过数据预处理和情感分析，使用 Gephi 软件绘制用户关系网络图。其次，构建视点演化模型，采用基于贝叶斯规则的视点更新公式计算几种不同发布中心网络关系内用户视点演化;结果表明：在舆情传播中，以媒体为中心的网络关系的观点演变方向趋于一致，这表明了媒体的观点引导在信息传播中的重要性。这些分析和技术为政府和媒体控制和引导网络舆情提供了一定的参考。