The Conceptual Framework Of Pakistani English Speech An Experimental Study Of Pakistani English For Speech Recognition And Machine Learning Modeling

Abstract

The primary objective of this research is to provide a comprehensive overview of the potential experimental and phonetic exponents of acoustical features of Pakistani English Speech. The Pakistani English research team is determined to document the acoustic structures of Pakistani English speech, which includes young Sindhi ESL learners’ Speech, Voice Onset Time (VOT), Lexical Prominence, Vowel Quality, Pitch, Intonation, and Prosodic Features for Speech Recognition, Machine Learning, and Mathematical Modeling in Pakistani English. The research study will be conducted by selecting seven pairs of disyllabic words following the methodology of Beckman (1986) and Fry (1955, 1958). These stimulus pairs will be formed from word forms such as contract, desert, object, permit, rebel, record, and subject. Each target word will be elicited in isolation and in a semantically neutral frame sentence "I said __ this time" and will be accompanied by associated context sentences created specifically for each word. The data will be collected from forty young Sindhi ESL learners, a diverse group of university undergraduate students, native speakers of Sindhi, Urdu, Punjabi, Pashto, and Balochi. The primary study will test the hypothesis that disyllabic words have either first-syllable stress in English nouns or second-syllable stress in English verbs. The initial pool of 100 students will be recruited for this study, resulting in a total of 130 participating students who speak four different L1 across the provinces of Pakistan. In the second study, 30 undergraduate Sindhi ESL learners will be recruited to record their voice samples in an anechoic chamber to examine whether Sindhi speakers transfer their L1 negative Voice Onset Time to L2 English-voiced stops. This study will result in a total of 1080 tokens for analysis. The study will generate algorithms through a Python Coding System for speech recognition and machine learning modeling based on the acoustic datasets. We are confident that our research will make significant contributions to the field of acoustic-phonetics in Pakistani English speech for speech recognition and machine learning fields and for teaching and learning Speech Science: Experimental Phonetics for BS/MS/PhD English/Linguistics/Computer Speech Scientists/researchers in the field.