Building MEDISCO: Indonesian Speech Corpus for Medical Domain
Muhammad Reza Qorib, Mirna Adriani · 2018
In this paper we report our work of building MEDISCO: Medical Indonesian Speech Corpus. The medical text corpus is collected from five Indonesian online medical consultation websites. From the text corpus, we created a speech corpus that consists of 360 sentences read by 13 speakers. In total, our speech corpus contains 731 medical terms and consists of 4,680 utterances with total duration 10 hours.