Word Clouds and Topic Model Mining Based on Statistical Analysis
Cheng Yu-shen · Journal of Anqing Teachers College · 2014
With the rapid development of internet and other information technologies,networks are accumulated with vast semistructured and unstructured text data. It will be a primary mission to statistical analysis workers that how to get the required information,and show it with an efficient and visual information graph from those massive electronic documents. Word clouds is a new text displaying way of information graph expressing. In the present work,we make some pretreatment of removing the number and the stop word in the text by a text mining method of word clouds and topic model. Then,we make Chinese word segmentation,build corpus and set up document-term matrix. Finally,we present the mining result with word clouds and topic model. The experiment statistically analyses the data of the rough set conference summaries using R language and make a contrast with word cloud generator of Tagxedo. These results indicate that the method of this paper has a better effect in mining and easy acquire important information from text,such as topic model.