Proposal of Enterprise Information Mining by Integrating Method to Extract Strategy- Words and Method to Extract Characteristic Corporate Financial Conditions

Abstract

We have developed the stratified N-Gram method, which is able to extract coined words and phrases those are difficult to extract by morphological analysis. Moreover, we proposed the method to extract the strategy-words with differences among enterprises and changes in time sequence, based upon stratified N-Gram. We also proposed the method to extract companies and periods, which are presumed that they had been under characteristic business conditions (e.g. unique management policy), by identifying financial indicators that show characteristic behavior against other competitors for specific period. In this article, we propose a concept of "Enterprise Information Mining", which extract valuable information from time sequences of both text data and corporate financial data.