基于朴素贝叶斯网络的微博话题追踪技术研究

Research on Micro-blog Topic Tracking Based on Naive Bayesian Network

根据微博的特点,提出了基于朴素贝叶斯网络模型的微博话题追踪算法,在改进型DF的文本特征选择方法的基础上,通过构建朴素贝叶斯网络模型,设计并实现对微博话题的追踪系统。实验结果表明,基于朴素贝叶斯网络的微博话题追踪系统具有分类简单、高效的优点,特别适合对微博热门话题进行追踪。

According to the characteristics of micro-blog,this paper proposes micro-blog topic tracking algorithm based on na.ive Bayesian network model,and at the base of the foregoing improved DF text feature selection method,the micro-blog topic track.ing system is designed and implemented. The experimental results show that micro-blog topic tracking system based on the naiveBayesian network model is simple,efficient,especially suitable for tracking hot topics on micro-blog.