Today, because there are many academic societies in the same academic fields, it is essential to differentiate the society from other similar ones. The features of the academic society is thought to be represented by papers published on the journal. However, it is difficult to recognize the differences of the features of societies by comparing manually the contents of large amounts of papers. This study picked up the academic societies related to the field of industrial engineering, and attempted to compare their features quantitatively by using text-mining. We developed a new method to demonstrate the feature of the societies more precisely than existing methods. As a result, we could obtain useful information for understanding the differentiation of academic societies.