Linguistic and Statistical Analysis of Audio Texts
Abstract
This article provides an in-depth analysis of the phonetic, morphological, and syntactic features of Uzbek audio texts based on linguostatistical analysis. The study integrates modern information and communication technologies, particularly natural language processing (NLP), speech technologies, and corpus linguistics methods. The article scientifically examines phonetic analysis (elision, assimilation, coarticulation), morphological modeling (word forms, affixes), and syntactic structures (asyndetic sentences, introductory words) conducted on audio texts. The Praat software was used for experimental analysis, while the uzbekcorpus.uz platform was utilized for statistical modeling. Based on the research results, practical recommendations are provided for creating automatic transcription, machine translation, ASR, and NLP systems for the Uzbek language. This article contributes to new scientific directions within the framework of integrating linguistics and artificial intelligence.