Mr. Jumpei Nishikawa
Dr. Junya Morita
The difficulties encountered by children during language development varies among individuals. In particular, immaturity in phonological awareness, which supports speech perception, results in various speech defects. Accordingly, it is important to estimate the individual mechanism behind these problems to ensure proper support. In this study, we propose a method for estimating individual defects in the phonological process using cognitive models. As a preliminary step to targeting phonological processing difficulties in real world, we conducted an experiment with native adult speakers. Audio filters were applied to the output of the system to simulate phonological difficulties. This initial feasibility study revealed consistency in model preferences among participants when a particular audio filter was used. We consider that this study provides an important step toward the realizations of individualized cognitive modeling for mitigating various difficulties in language acquisition.