Proposed Arabic word lists for subjective speech intelligibility tests via a communication system

Abstract

This study addresses the critical gap in speech intelligibility testing for Arabic-speaking regions by introducing the first standardized, phonetically balanced Arabic word lists, aligned with ANSI S3.2. We curated 20-word lists, totalling 1,000 words, selected for their phonetic representation and frequency in the Modern Standard Arabic. These words were refined from an initial pool of 1,200 words through systematic evaluations of difficulty and familiarity. A key innovation here is the development of the Word Selector Application (WSA), a Python-based tool that facilitates randomized word selection, ensuring diversity and phonetic balance. The WSA significantly contributes to standardizing speech intelligibility tests in Arabic, enhancing the accuracy and reliability of assessments in many acoustic disciplines. Not only does this work pave the way for inclusion in international standards but it also underscores the need for comprehensive Arabic standards in subjective speech intelligibility testing, marking foundational steps towards bridging linguistic disparities in acoustic assessment.