Implementasi Speech to Text untuk Mempermudah Catatan Praktik Diagnosis Pasien Dengan Metode NLP

Ridho Fajar Fahturohman (Universitas Pembangunan Nasional “Veteran” Jawa Timur)
Nafis Pratama Putra (Universitas Pembangunan Nasional “Veteran” Jawa Timur)
Panggih Santri (Universitas Pembangunan Nasional “Veteran” Jawa Timur)
Anggraini Puspita Sari (Universitas Pembangunan Nasional “Veteran” Jawa Timur)

Journal of Multidisciplinary Inquiry in Science, Technology and Educational Research
Vol. 1 No. 3c (2024): JULI (Tambahan)

Medical diagnoses in students' field study practice activities are very important, as data in preparing field study practice reports. The large amount of diagnostic data that has to be retyped manually makes time ineffective. Based on this problem, this research implements Speech to Text (STT), a feature that can convert voice into text using the Natural Language Processing (NLP) method using the Python programming language. The NLP method is used in this research because its role is to interpret a language, both written and spoken, so it really supports the operation of STT. Students just need to record the diagnostic conversation with the patient, then the .wav format file is processed in a system that will provide written output, so students just have to copy it into a report. The test used some audio files that had different durations and noise levels. The system can convert voice into text with a success percentage of 93,34%.