Stability of Textual Information Analysis and Interpretation of Evaluation Reports

Abstract

Data analysis methods for textual information are popular research techniques which deepen global understanding on characteristics of accumulated text data such as evaluation reports, and also have possibility leads to new knowledge discovery. We focus on text information concerning evaluation reports of National University Corporation Evaluation. We conducted research on information extracting and text analyzing for grasping the global and local characteristics of the report data. Moreover, we considered stability of analysis in case of data perturbation.