Study on Improving Machine Learning Discriminators using Vocal Parameter of Korean Learners

초록

South Korea has transformed from one of the world's poorest countries into one of its wealthiest. Since the Korean War, the nation has not only elevated its standard of living through technological innovations but has also become a prolific producer of globally popular cultural content. This rise in the popularity of K-culture has attracted learners from various countries to the Korean language. Located strategically between China and Japan, Korea draws numerous foreign language learners, including international students and industrial trainees from countries such as Vietnam and Uzbekistan. Pronouncing Korean accurately poses challenges due to the pronunciation habits rooted in the learners' native languages. Previous research focused on analyzing the pronunciation characteristics of Chinese or Vietnamese speakers and proposed the use of a Support Vector Machine (SVM) discriminator. This study aims to refine the parameters of the SVM's hyperplane to better distinguish pronunciation variations. It introduced research that leverages this discriminator to facilitate more precise Korean pronunciation among non-native speakers.

키워드