In the development of an automatic speech recognition (ASR) system, anextensive study of the basic phonetic alphabet is performed to collectinformationregarding phonology and phonetics of the language or dialect in question(modemstandard Arabic or MSA in this case). In addition, terminological andtranscriptionalproblems are identified with respect to the language or dialect in question.Next, basedon feature description (rather than symbol shapes), the symbols in theliterature aremapped to a single or more recent phonetic alphabet. Lastly, from a maximalsetcontaining all the phonemes, allophones, and transliteration symbols, areduced set iscreated with a compact set of phonetic alphabets. Memory consumption isgreatlyreduced in a computer system by using this compact set of phonetic alphabets
展开▼