Abstract—This document outlines the use of an intelligent speech therapy system as a learning aide for individual having a speech disorder known as Velopharyngeal Insufficiency. The goal is to detect speech disorder under a phoneme specific etiology such as plosives, affricates, and fricatives. The system used speech recognition to capture and evaluate speech signals and give an accurate and correct module for the user. The digital signals were processed, analyzed and computed using Mel Frequency Cepstral Coefficient (MFCC). Through these methods, system will come up with a speech evaluation results and modules for therapy. The system was tested by random students having an accuracy of 85.55%.

Index Terms—MFCC, velopharyngeal insufficiency, phonetically-balanced words, silence removal, pre-emphasis, framing, humming, windowing, fast fourier transform (FFT), mel filter bank.

J. C. Castillo, J. Z. Fabian, M. L. Magbanua, and J. R. Saloma are with the Department of Computer Science, Polytechnic University of the Philippines in Sta. Mesa, Manila 1008 Philippines (e-mail: castillojoanne0208@yahoo.com; joseph_fabian@yahoo.com; markmagbanua@rocketmail.com; jrhio08@yahoo.com).
R. A. Sagum is with the Department of Computer Science, College of Computer and Information Sciences, Polytechnic University of the Philippines in Sta. Mesa, Manila and is also with the Faculty of Engineering, University of Santo Tomas in Manila, Philippines (e-mail: rasagum@pup.edu.ph).

Cite:Joanne C. Castillo, Joseph Z. Fabian, Mark Lean V. Magbanua, John Rhio F. Saloma, and Ria Sagum, "TheraSpeech: An Intelligent Speech Therapy System for Velopharyngeal Insufficiency," International Journal of Computer and Communication Engineering vol. 4, no. 1, pp. 22-25, 2015.