Arabic Automatic Speech Recognition: A Systematic Literature Review

Automatic Speech Recognition (ASR), also known as Speech-To-Text (STT) or computer speech recognition, has been an active field of research recently. This study aims to chart this by performing a Systematic Literature Review (SLR) give insight into the ASR studies proposed, especially for Arabic language. The purpose is highlight trends about and guide researchers with most significant published over ten years from 2011 2021. SLR attempts tackle seven specific questions related toolkits used developing evaluating ASR, supported type language, feature extraction/classification techniques, performance existing gaps facing researchers, along some future research. Across five databases, 38 met our defined inclusion criteria. Our results showed different open-source support recognition. prominent ones were KALDI, HTK, then CMU Sphinx toolkits. A total 89.47% retained cover modern standard Arabic, whereas 26.32% them dedicated dialects Arabic. MFCC HMM presented extraction classification respectively: 63% papers based on 21% HMM. review shows that systems depends mainly criteria availability resources, techniques acoustic modeling, datasets.