An Efficient Method for Finding Improvements in Japanese Management Training Programs Using Text Mining

Abstract

While changes in the business environment are accelerating, working life is lengthening as we enter an era of 100-year lifespans, making it necessary for working adults to continually improve their skills. Training for working professionals is becoming increasingly important, and educational institutions continuously strive to improve their training programs. However, it is difficult to efficiently extract important information from the vast number of participant satisfaction survey comments for the purpose of program improvement. In this study, we used text mining to examine how to efficiently retrieve important information from comment data of satisfaction surveys on management training programs in Japan. In general, a method for extracting dissatisfaction from the polarity of words in comments has been considered. Nonetheless, the analysis results confirmed that even when describing dissatisfaction or demands expressions, Japanese managers do not use straightforward expressions, but rather characteristic expressions. Based on an understanding of their politeness and considerate expressions, we learned that extracting and utilizing keywords from the comments of a small group was effective in efficiently detecting the overall dissatisfaction and demand expressions. It was also discovered that the same working adults used different words to express dissatisfaction and demands, depending on their job positions. Using these novel study findings, essential improvements can be efficiently explored from post-training satisfaction questionnaires, contributing to improving the quality of education for working adults in the future.