According to the characteristics of micro-blog text,this paper takes the micro-blog three kinds of special symbols,"@","//"and"#",into consideration for topic extraction analysis of micro-blog text. On the basis of the traditional LDA model,a micro-blog tag LDA model is established,which considers the micro-blog theme,forwarded micro-blog,micro-blog comment con?tent,and enhances the accuracy of the extraction of micro blog theme. Experimental results show that the algorithm proposed in this paper has a good effect on the topic extraction of Sina micro-blog.%论文根据微博文本所具有的特点,将微博文本具有的三种的特殊符号:"@"、"//"和"#"纳入微博主题提取的分析中.在传统的LDA模型的基础上,建立一种微博标签的LDA模型,该模型考虑到微博的主题、转发的微博、微博的评论等内容,增强对微博主题的提取的准确性.实验结果表明,论文提出的算法对于新浪微博的主题提取的效果良好.
展开▼
