Setting Up A Complex Model Of Speech Analysis: Pilot Study Of Late Bilingual Speech

Abstract

Modern applied linguistics includes a substantial amount of spoken language analysis, and current Natural Language Processing techniques allow working with large amounts of audio data. This article demonstrates a method for conducting primary research of late bilinguals' speech, which is particularly relevant in the context of modern globalization. Using a small sample as an example, the paper presents the methodology testing the informants' speech, including technical approaches for speech collecting, processing, and interpretation. The dataset for the analysis is the interview recordings, which took place after the informants watched a silent film.