Intelligence system design for classification of regional dialects / Prof. Ir Dr. Shah Rizam Mohd Shah Baki, Prof. Dr. Mohd Nasir Taib and Mohd Azri Abd Aziz

With the advent of notebook and hand held computers in the consumer market, dialect-input system will gain prominence in the future over keyboard and penbased system. This report presents an intelligent dialect classifier using Waveletbased Fuzzy system: The use of wavelet to capture important time and frequency feature for the speech. The classification for the Malaysian regional Dialect is achieved by using fuzzy set and their associated membership functions. The overall performance of the wavelet-based Fuzzy system is able to classify the regional dialect. The simple dialect classification can be implement with a personal computer and once a limited system has been implemented, using faster computer and specialized hardware as necessary can scale its capabilities.