A Visualized Pattern Discovery Model for Text Mining Based on TF-IDF Weight Method

This paper focuses on the pattern discovery of text and builds a visualized pattern discovery model base on TF-IDF Weight Method. The model can discover patterns in the form of hotspots, cross-distribution and development trend. With the help of relevant tools and methods the model realizes the pattern discovery in the way of visualization which has a better human-computer interaction performance. It is testified by the research related text data of the pattern discovery outcome. The results show that the model could well express the hotpot directions, cross-distribution situations and development trends. The model provides a possibility of pattern discovery especially in a visualization method.