Computerized Method for Diagnosing and Analyzing Speech Articulation Disorder for Malay Language Speakers

Abstract

This study aims to develop a computerized technique that uses speech recognition as a helping tool in speech therapy diagnosis for early detection. Somehow speech disorder can be divided into few categories which not all type will be fully covered in this research. This study only purposes a solving method for diagnosis of a patient that suffers from articulation disorder. Therefore a set of Malay language vocabulary has already been designed to tackle this issue where it will cover selected Malay consonants as a target words. Ten Malay target words had been choose to test the recognition accuracy of this system and the sample are taken from real patient from Hospital Sultanah Aminah (HSA: Speech therapist at Speech Therapy Center) where the hospital assists in the clinical trial. The result accuracy of the systems will help the Speech Therapist (ST) to give an early diagnosis analysis for the patient before next step can be purposed. An early percentage of correct sample achieved almost 50% in this experiment.