Answers are grounded in this paper and cite their sources.
Phonological features in Automatic Speech Recognition (ASR) has been found from people experiencing the interlanguage. Numerous studies have used ASR as a tool to convert speech form into text form. Initially, it extracts features and elements of a speech, including the phonological features. The negative transfer is a phenomenon caused by L2 learners try to, intentionally or not, bring their respective L1 regularities into the L2. Furthermore, it creates a new language in the process, called the interlanguage. One can found the negative transfer phenomenon in the interlanguage, which is commonly experienced by L2 learners. This phenomenon will be most likely to be examined in its phonological features of respective first language (L1). In this paper, we investigate the negative transfer in the phonological features from different ASRs using different models, techniques, and algorithms. In addition, we also present the respective language phonology model to highlight the why factor causing the negative transfer phenomenon. This paper is a literature review regarding the effort of phonological features extraction for Automatic Speech Recognition (ASR) in terms of interlanguage.
No takes yet. Share an insight, caveat, or question.
Adi et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: