In this paper, we describe a method for extracting evaluations and desires from the free-description results of a survey. For this purpose, we employ a text-mining system and a questionnaire, and we present the extraction results. Using a questionnaire, we conduct a web-based survey after each session of computer programming class, in order to enhance the effectiveness of the next session. Free-description observations are important for extracting the students’ desires. However, reading the free-description responses requires considerable time. Moreover, computer-based analysis of these responses is difficult. Therefore, we propose a text- mining system based on natural language processing, which enables us to automatically classify sentences for class evaluation and for the determination of student motivation and opinions. The proposed system facilitates our understanding of students’ evaluations and desires ; such information can enable us to conduct classes more effectively.
