基于深度学习的微博舆情监测模型研究与实现

Research and implementation of Weibo public opinion monitoring model based on deep learning

为了实时监测和分析新浪微博上的舆论情况,建立一种基于深度学习的微博舆情监测模型。提出了基于Java的分布式数据爬取框架和基于Elasticsearch的分布式搜索存储方法,有效地提升了舆情监测模型的性能。提出了融合改进注意力机制的Bi-LSTM情感分析方法和基于情感分析的舆情预警等级计算方法,很好地实现了对微博热搜话题的实时舆情监测。

In order to monitor and analyze the public opinion situation on Sina Weibo in real-time,a Weibo public opinion monitoring model based on deep learning is established.It proposes a distributed data crawling framework based on Java and a distributed search storage method based on Elasticsearch,effectively improving the performance of the public opinion monitoring model.A Bi-LSTM sentiment analysis method that integrates improved attention mechanism and a public opinion warning level calculation method based on sentiment analysis are proposed,effectively achieving real-time public opinion monitoring of hot topics on Weibo.