Graph Discovery and Visualization from Textual Data

摘要: This chapter tackles the problem of knowledge discovery in text collections and dynamic display discovered knowledge. We claim that these two problems are deeply interleaved, should be considered together. The contribution this is fourfold: (1) description properties needed for a high-level representation concept relations text; (2) stochastic measure fast evaluation dependencies between concepts; (3) visualization algorithm to structures; (4) deep integration visualization, i.e., placement nodes edges automatically guide displayed. resulting program has been tested using specific data sets based on domains molecular biology WWW how-tos. have also integrated proposed methods more refined Web mining process.