2021 年 23 巻 4 号 p. 397-406
With the globalization of the field of activity, it is an ongoing challenge to improve the English language skills of engineers and researchers. In this study, we propose a method for estimating English language ability using self-organizing maps with the gaze and EEG while reading English. 26 participants with different TOEIC L&R Test scores were used as the samples, and the validity was tested with 8 new participants. The participants read English texts on a PC display for one minute and rested for one minute with the eyes closed, meanwhile their gaze and EEG were recorded. In terms of gaze information, we analyzed pauses, regressions, reading speed, pupil diameter, and blinks during English reading. For the EEG, we focused on the frequency bands classified as the α wave, analyzed them with the frequency spectrum area. The characteristics of each cluster on the self-organizing map were analyzed by focusing on the listening score and reading score of the TOEIC L&R Test as the major strengths and weaknesses in English language proficiency. It was concluded that it is possible to estimate the strengths and weaknesses from the clusters classified on the self-organizing map and to suggest learning methods to improve the scores.