Artificial intelligence in language education: introduction of Readizy

Abstract

Using artificial intelligence technology, we proposed Readizy to help reading teachers and readers with reading diagnosis. Readizy is an interdisciplinary program based on theoretical tenets and technological developments in linguistics, psychology, and affective computing. This study is part of our overall program; we plan to extend the program from a technological perspective. Specifically, through facial expression recognition techniques, we used Readizy to document readers’ facial expressions during the reading process and categorize the expressions into seven classes. We conducted experiments and compared our experimental results with those in the existing FER2013 and CK+ systems. To create expression classifications specifically for readers, we made Readizy unique and powerful for tracking emotional changes with high classification accuracy. This study contributes to affective computing and reading education.