LITERATURE REVIEW

 Based on the review of previous studies, this study will focus on two areas.

 

1. Automatic Speech Recognition (ASR) for Pronunciation Learning.

Automatic speech recognition (ASR) is a machine-based, autonomous method for decoding and recording spoken language (Levis & Suvorov, 2012). Distinguishing between speech recognition and speech understanding (or speech identification) is essential, where speech understanding is the process of interpreting the meaning of an expression rather than simply transcribing the words. Speech recognition is different from voice recognition. Speech recognition is related to a machine's ability to identify spoken words (i.e., what is being said). In contrast, speech recognition is associated with a machine's ability to recognize a person's speaking style (i.e., who is speaking the words). According to (Levis & Suvorov, 2012), Speech recognition systems can be distinguished based on three main things: speaker dependency, speech continuity, and vocabulary size. Based on the speech data used for training, ASR systems can be speaker-dependent (where the system needs to be trained for each speaker), speaker-independent (where the system can recognize new speakers because its training data comes from many speakers), or adaptive (where the system starts without recognizing a particular speaker, but then adapts to a specific speaker through training). Based on speech continuity, there are (a) isolated word recognition systems, which recognize words spoken fragmentarily; (b) connected word recognition systems, which can recognize words spoken without pauses; (c) continuous speech recognition systems, which can recognize complete sentences without pauses between words; and (d) word scanning systems, which retrieve words or phrases from continuous speech. Finally, ASR systems can also be differentiated based on the size of the vocabulary used in training, whether small or large.

Automatic Speech Recognition (ASR) technology has been suggested in recent studies on technological developments as a valuable tool to enhance learners' FL speaking skills. Cucchiarini et al. (2009) found that the feedback provided was compelling enough to help students enhance their pronunciation after practice in a relatively short period, even though the developed ASR system has not achieved perfect accuracy in detecting students' pronunciation errors. Research by McAndrews (2020) supports the use of technology in language learning, as it has proven effective in improving students' pronunciation skills, primarily in prosody. Researchers have started using ASR technologies that convert speech into text freely, such as Google Speech Recognition (GSR) (e.g., Mroz, 2018) for spoken exercises to improve pronunciation skills. Previous research has generally found that learners have a favorable view of pronunciation training using ASR, as this technology allows them to practice independently (McCrocklin, 2016) and provides direct feedback (Inceoglu, 2022; Liakin et al., 2017).

 

2. Teaching Pronunciation

Many aspects must be considered when speaking, one of them being pronunciation. Pronunciation has an essential role in communication, as when someone does not pronounce words clearly, it will undoubtedly cause the meaning to be conveyed, not delivered. Good pronunciation is necessary for effective communication, although other language skills are also essential (Alghazo, 2015). Teaching pronunciation holds significant importance as it directly impacts the interpretation of sentences, typically conveyed through both segmental and suprasegmental elements. Segmental features pertain to the smallest phonetic sound units (Pennington & Richards, 1986, p. 208). Proper pronunciation teaching is essential to reduce misunderstandings in communication, especially in foreign languages. Thus, pronunciation teaching should focus on helping students develop clear and understandable speech (Alghazo et al., 2023). For several reasons, teaching proper pronunciation is essential in communication, especially in the scope of foreign languages. First, pronunciation plays a significant role in achieving effective communication. Adequate pronunciation can help ensure that the recipient of the message can understand the sentence we convey because if there are errors in pronunciation, it can hinder the communication process because of mistakes in interpreting meaning. Secondly, proper pronunciation also affects someone's listening ability. Learners who pronounce and recognize sounds correctly find it easier to understand the original speaker. Third, adequate pronunciation indicates that someone has good language skills and can help increase learners' confidence. (Alsuhaibani et al., 2024). Pronunciation teaching has a very significant role in improving the accuracy of language production among L2/FL learners of various languages. (Piske et al., 2001).