Pronunciation Feedback Research Articles

AI-powered language learning: The role of NLP in grammar, spelling, and pronunciation feedback

Abstract. The introduction of artificial intelligence (AI) and natural language processing (NLP) in language learning has transformed the practices of education. This paper analyzes the role of AI-powered tools in English language learning with specific focus on automated essay scoring (AES), speech recognition, and advanced grammar and spelling checkers. By examining the accuracy of the most effective AES systems, it is evident that automated feedback has increased the quality of writing by providing detailed, consistent, and objective feedback on submitted essays. Speech recognition is another tool that is being used to assess pronunciation and fluency of learners. The software is constantly improving, with advances in NLP and machine learning. The technology for fluency tests is becoming more accurate, and is helping to improve learner outcomes significantly. Lastly, we examine advanced grammar and spelling checkers, which are increasingly used for correcting the contextual errors of individual learners. The AI systems adapt to the user and their needs, which make the corrections more effective. This paper offers an overview of how AI is leading to personalized and effective language learning, ultimately enhancing the educational experience of students and embracing future possibilities in education technology.

Quantifying the impact of ASR-based instruction: What does the iSpraak platform learner data show?

Computer-assisted Pronunciation Training (CAPT) tools have become increasingly dependent on Automatic Speech Recognition (ASR) technology to provide automated corrective pronunciation feedback to learners. The extent to which ASR-based tools measurably improve second language (L2) pronunciation is of great interest to language educators globally, and Computer-assisted Language Learning (CALL) researchers. Studies to date have largely been conducted by research practitioners with small-to-medium sized samples at single institutions. The findings and conclusions drawn from such small-scale data collection might be significantly bolstered by analysing the vast stores of learner data from large CAPT platforms. This study is informed by a sizable eight-year dataset from iSpraak, an open-source pronunciation tool designed to model and evaluate L2 speech. Quantitative analysis of anonymised learner interactions with this application reveals significant gains in intelligibility measures across multiple languages. Results also suggest that the extent of ASR’s ability to improve learner pronunciation may be L2 dependent.

2D Spectrogram analysis using vision transformer to detect mispronounced Arabic utterances for children

Pronunciation feedback is essential for teaching languages to children, urging the need to create computer-assisted pronunciation training (CAPT) systems to automate this process. Most of the current CAPT systems for Arabic Language examined a wide age range, including children, with a dominant adults’ portion. However, these systems did not consider many challenges facing children’s voice, in terms of the acoustic difference between children and adults. Moreover, due to the lack of publicly Arabic datasets, Arabic-related systems were evaluated on datasets consisting of the sound of Arabic letters, with minor efforts to contain words or sentences. In this paper, we propose the Arabic Utterance Mispronunciation Detector for Children (AUMD-Child) system that detects the mispronunciation of children for Arabic words at real-time. The proposed system is based on the fusion of Vision Transformer (ViT) and transfer learning-based model (AlexNet). A novel public dataset is constructed, in which 16 words have been collected from children aged between 7 and 12 years old to evaluate our proposed system. The experimental results, compared to the state-of-the-art models in literature, showed that the accuracy of our proposed AUMD-Child system exceeded the handcrafted features-based models by an average of 7 %, the CNN features-based models using AlexNet and SVM by 5 %, transfer learning-based model using AlexNet by 4 %, and transformers-based model using ViT by 2 %, achieving an average accuracy of 91.81 % for detecting the mispronunciation patterns in Arabic words but in an average processing time of 33 ms, exceeding the other state-of-the-art models with 13 %, 0.03 %, 0.06 %, and 23 % respectively.

Enhancing self-regulation and learner engagement in L2 speaking: exploring the potential of intelligent personal assistants within a learning-oriented feedback framework

IntroductionDeveloping L2 speaking proficiency can be challenging for learners, particularly when it comes to fostering self-regulation and maintaining engagement. Intelligent Personal Assistants (IPAs) offer a potential solution by providing accessible, interactive language learning opportunities.MethodsThis mixed-methods study investigated the effectiveness of using Google Assistant within a learning-oriented feedback (LOA) framework to enhance L2 speaking proficiency, self-regulation, and learner engagement among 54 university-level EFL learners in China. Convenience sampling assigned participants to either an experimental group (n = 27) using Google Assistant with tailored activities or a control group (n = 27) using traditional methods. The Oral Proficiency Interview (OPI) assessed speaking performance. Self-reported questionnaires measured L2 motivation and the Scale of Strategic Self-Regulation for Speaking English as a Foreign Language (S2RS-EFL) evaluated speaking self-regulation. Additionally, semi-structured interviews with a subsample of the experimental group provided qualitative insights.ResultsThe Google Assistant group demonstrated a statistically significant improvement in speaking performance compared to the control group. While no significant difference in motivation was found, thematic analysis of interviews revealed perceived benefits of Google Assistant, including increased accessibility, interactivity, and immediate pronunciation feedback. These features likely contributed to a more engaging learning experience, potentially fostering self-regulation development in line with the core principles of LOA.ConclusionThis study suggests Google Assistant as a promising supplementary tool for enhancing L2 speaking proficiency, learner autonomy, and potentially self-regulation within an LOA framework. Further research is needed to explore its impact on motivation and optimize engagement strategies.

Lukio-opiskelijoiden käsityksiä vieraan kielen ääntämisen opetuksessa saadusta palautteesta

In foreign language teaching, feedback is an essential part of learning pronunciation. It raises the learner’s awareness of the strengths and weaknesses of their pronunciation. Yet previous research has shown that teachers often feel uncomfortable when giving feedback on learners’ pronunciation. We conducted a survey among Finnish high school students to find out their pronunciation targets and what kind of feedback they would like to receive. The survey revealed that a clear majority (80 %, n=1953) of respondents appreciated good pronunciation and wished for native-like pronunciation. They hoped to receive feedback from the teacher, and they understood the importance of corrective feedback for learning. Based on our findings we encourage teachers to provide positive and corrective pronunciation feedback to their students to promote effective learning.

Vlog Task Accompanied with Individualized Pronunciation and Grammar Feedbacks Improve Speaking Ability

Speaking plays both roles as process and objective of learning in language classroom aimed at preparing future communication competence of students. Using Vlog as speaking activities to demonstrate ideas and knowledge can be recreational for the millennial generation. However, grammatical forms and pronunciation accuracy are often neglected areas in such tasks, hence students’ grammatical competence and pronunciation accuracy are not as well developed as their fluency in speaking. The current study innovated to combine Vlog-speaking activities with individualized pronunciation and grammar feedbacks from the teacher in order to not just elevate the millennial students’ Vlog talk performance, but also increase wellness of grammatical forms, accuracy of pronunciation, and clarity of content, hence boost higher confidence. The study measured the effectiveness of this holistic instruction in improving students’ English public speaking skill, and investigated how the students perceived the Vlog task, the individualized grammar feedback, and pronunciation feedback from the teacher. Sixty-two students of health from two equal intact classes at Politeknik Kesehatan Malang were involved as experimental and control groups in the quasi-experiment study. The result of statistical analysis showed that the speech performance of the experimental students was significantly higher than the conventional lecture-treatment control group. In addition, the experimental students’ products of speech seemed to be very creative, fun, and attractive. The results also showed that students had very positive perspectives towards Vlog task although challenging because they felt it could develop their English communication skills

Speakometer: English Pronunciation Coach

Consonant and vowel sounds of English (segmentals) carry a significant weight in communication. Pronunciation instruction focusing on segmental features has been found to be highly effective (e.g., Thomson & Derwing, 2015). However, students with different first languages (L1) or even students from the same L1 backgrounds, have different pronunciation needs. With limited class time, teachers cannot be expected to cater to the pronunciation needs of every student. This has made individualized pronunciation instruction, which enables pronunciation instruction tailored for the needs of each second language (L2) learner, a requirement in today’s language classrooms (Levis, 2007). The growing number of computer-assisted pronunciation training (CAPT) tools have been responding to this need, making individualized pronunciation instruction, as well as individualized feedback more feasible and available for L2 speakers. Chun (2012) asserts that in order for a CAPT tool to be effective, it must contain “auditory and visualization features, automatic speech recognition (ASR), and appropriate and accurate feedback” (p. 8). Speakometer, an online application that provides segmental practice for its users, was built around Chun’s (2012) three pillars, with a strong auditory feature combined with an ASR to provide learners with relevant pronunciation feedback. The application uses an artificial intelligence (AI) algorithm and ASR to rate the user’s spoken English pronunciation. It is targeted for all users who aim to improve their English pronunciation. The users are provided with immediate feedback, which appears on the screen as verbal (e.g., “Very good”), along with the image of a ‘speakometer’ displaying four colors for the rating: red, orange, yellow and green.

Enhancing the auditory usage of L2 e-storybooks by precedingly providing learners with perceptual intervention

ABSTRACT E-storybooks are the most important auxiliary learning materials for English-as-Foreign-Language (EFL) learners. Few studies reported positive results on pronunciation and speech development from using audio accompanied e-storybooks, although audios are expected to play the role of recasts in providing pronunciation feedback for learners. Presumably, it is learners’ insufficient perceptual ability that prevents them from replicating and simulating native-like speaking. Thus, based on the perception-precedes-production hypothesis, this study investigates the possibility of enhancing learners’ oral repetition of the reading attached to an e-storybook by precedingly providing learners with reading-while-listening-based perceptual intervention. The experimental results show that the intervention had a significant effect on the long-sentence repeating rate, and minor effects on phoneme recognition and memorisation. Moreover, the eye-gaze patterns of the participants in the experimental group reveal that the synchronisation of the visual- and auditory attention in reading-while-listening-based perceptual training is vital for the intervention efficacy.

Affordances and Constraints of Intelligent Personal Assistants for Second-Language Learning

This paper provides a research synthesis of intelligent personal assistants (IPAs) – that is, cloud-based virtual assistants such as Alexa, Google Assistant, and Siri – for second language (L2) learning. The article also offers a theoretical justification for the use of IPAs in language learning and outlines the affordances and constraints of the technology. Finally, the article proposes directions for future research on the topic of IPAs based on the aforementioned review of L2 literature. While research indicates that IPAs increase opportunities for meaningful communication in the target language, enhance speaking ability, and provide indirect pronunciation feedback, they also present some challenges for L2 learners – namely, IPAs struggle to reliably understand L2 speech, which may limit the usability of virtual assistants among heavily accented learners.

A computer-aided speech analytics approach for pronunciation feedback using deep feature clustering

A computer-aided speech analytics approach for pronunciation feedback using deep feature clustering

A Longitudinal Evaluation of Tablet-Based Child Speech Therapy with Apraxia World

Digital games can make speech therapy exercises more enjoyable for children and increase their motivation during therapy. However, many such games developed to date have not been designed for long-term use. To address this issue, we developed Apraxia World, a speech therapy game specifically intended to be played over extended periods. In this study, we examined pronunciation improvements, child engagement over time, and caregiver and automated pronunciation evaluation accuracy while using our game over a multi-month period. Ten children played Apraxia World at home during two counterbalanced 4-week treatment blocks separated by a 2-week break. In one treatment phase, children received pronunciation feedback from caregivers and in the other treatment phase, utterances were evaluated with an automated framework built into the game. We found that children made therapeutically significant speech improvements while using Apraxia World, and that the game successfully increased engagement during speech therapy practice. Additionally, in offline mispronunciation detection tests, our automated pronunciation evaluation framework outperformed a traditional method based on goodness of pronunciation scoring. Our results suggest that this type of speech therapy game is a valid complement to traditional home practice.

English speech sound improvement system based on deep learning from signal processing to semantic recognition

With the global integration and the increasing level of China's internationalization, the demand of Chinese people for English learning is growing rapidly. At the same time, English has gradually become one of the most frequently used languages in the world economic exchanges and cultural information exchanges. After the material life demands of the members of the domestic society are fully satisfied, the learning and explanation of the corresponding knowledge content of the English language are gradually highly valued by the staff in various production fields. After AI proposed deep learning models, researchers also began to improve these models and use them in speech recognition. Through multi-layer nonlinear structure, the essential features are preserved and more abstract advanced features are extracted, so the recognition rate is improved. Computer-aided language learning will change the existing language teaching mode and learning environment, so that learners can learn independently anytime and anywhere. Deep learning can give learners accurate, objective and timely pronunciation evaluation and feedback guidance. It can also help learners find out the differences between their pronunciation and standard pronunciation through repeated listening and comparison, correct their pronunciation errors, and improve the efficiency of language learning. This paper constructs an English language improvement system based on deep learning. The experimental results show that the proposed method can effectively analyze the input speech signals and make corresponding feedback.

Using Intelligent Personal Assistants for Second Language Learning: A Case Study of Alexa

The proliferation of smartphones has given rise to intelligent personal assistants (IPAs), software that helps users accomplish day‐to‐day tasks. However, little is known about IPAs in the context of second language (L2) learning. Therefore, the primary objectives of this case study were twofold: to assess the ability of Amazon's IPA, Alexa, to understand L2 English utterances and to investigate student opinions of the IPA. Four university students of English as a foreign language (EFL) in Japan participated in the study, which involved each participant interacting with Alexa in a 20‐min session. Three sets of data were collected and analyzed to achieve the study's aims: learner‐generated command performance, interactive storytelling performance, and interviews. The quantitative results showed that Alexa accurately understood only 50% of learner commands, whereas comprehensibility during the interactive storytelling skill, Earplay, was much higher (90%). Three themes were identified from the interviews based on criteria developed by Hubbard (): hindered learner efficiency due to the lack of first language (L1) support, improved learner effectiveness through indirect pronunciation feedback, and better access to conversational opportunities. These findings demonstrate that EFL learners perceive Alexa to be a potentially useful tool to enhance language learning and underscore the need for additional (L2) research of IPAs.

Pronunciation Technology: Global Community and Innovative Tools in Forvo and NetProF Pronunciation Feedback

Pronunciation Technology: Global Community and Innovative Tools in Forvo and NetProF Pronunciation Feedback

Learner perceptions of online peer pronunciation feedback through P-Check

Receiving adequate pronunciation feedback is an ongoing challenge for l2 learners. Although instructors are the most important source of corrective pronunciation feedback (Szpyra, 2014; Timson, 2007), l2 learners can also benefit from peer pronunciation feedback (Lord, 2008; Kim & Yoon, 2014; Roccamo, 2015). This paper examines Japanese university students’ perceptions of online peer feedback on intelligibility (Derwing & Munro, 2005). To aggregate the feedback, P-Check, a plug-in for the Glexa lms (Version2, 2015) was developed and its recognition and dialog functions were piloted in two studies. In the first exploratory study (N = 59), a post-treatment survey and diaries were used to explore participants’ attitudes. Participants believed that peer feedback through P-Check was useful and had increased their awareness of their pronunciation strengths and weaknesses. In the second quasi-experimental study (N = 55), no significant difference was found in participants’ perceptions of the usefulness of peer feedback, whether delivered face-to-face or via P-Check. Recommendations for further development of P-Check and other similar plug-ins are given.

Learner perceptions of online peer pronunciation feedback through P-Check

Learner perceptions of online peer pronunciation feedback through P-Check

The pedagogical use of mobile speech synthesis (TTS): focus on French liaison

ABSTRACTWe examine the impact of the pedagogical use of mobile TTS on the L2 acquisition of French liaison, a process by which a word-final consonant is pronounced at the beginning of the following word if the latter is vowel-initial (e.g. peti/t.a/mi = > peti[ta]mi ‘boyfriend’). The study compares three groups of L2 French students learning how to produce liaison over a two-month period, following a pretest-posttests design within a mixed-methods approach to data collection and analysis. Participants were divided into three groups: (1) the TTS Group used a TTS application on their mobile devices to complete weekly pronunciation tasks consisting of noticing, listen-and-categorize, and listen-and-repeat; (2) the Non-TTS Group completed the same weekly pronunciation tasks in weekly sessions with a teacher; finally, (3) the Control Group participated in weekly meetings ‘to practice their conversation skills’ with a teacher, who provided no pronunciation feedback. The results indicate that, although all three groups improved in liaison production, if considered separately (within groups), only the two experimental groups improved over time. The discussion of our findings highlights the pedagogical use of mobile TTS technology to complement and enhance the teaching of L2 pronunciation.

Learning L2 pronunciation with a mobile speech recognizer: French /y/

This study investigates the acquisition of the L2 French vowel /y/ in a mobile-assisted learning environment, via the use of automatic speech recognition (ASR). Particularly, it addresses the question of whether ASR-based pronunciation instruction using a mobile device can improve the production and perception of French /y/. Forty-two elementary French students participated in an experimental study in which they were assigned to one of three groups: (1) the ASR Group, which used an ASR application on their mobile devices to complete weekly pronunciation activities, with immediate written visual (textual) feedback provided by the software and no human interaction; (2) the Non-ASR Group, which completed the same weekly pronunciation activities in individual weekly sessions but with a teacher who provided immediate oral feedback using recasts and repetitions; and finally, (3) the Control Group, which participated in weekly individual meetings ‘to practice their conversation skills’ with a teacher who provided no pronunciation feedback. The study followed a pretest/posttest design. According to the results of the dependent samples t-tests, only the ASR group improved significantly from pretest to posttest (p < 0.001), and none of the groups improved in perception. The overall success of the ASR group on the production measures suggests that this type of learning environment is propitious for the development of segmental features such as /y/ in L2 French.

Design of English Pronunciation Training System for Android

This paper is devoted to an English Pronunciation training system on the Android platform with functions like English transcription, word pronunciation demonstration, listening and repeating, pronunciation comparison, grading, pronunciation feedback and correction. The training system is designed and developed in line with the characteristics of Android operation system, giving full play to the advantages of Android platform by calling the relevant functions of function interface and components to improve the development efficiency of the system. The system is of high real-time and reliability with a series of multimedia techniques like the pronunciation mouth video-based animation, internal mouth structure figure, pronunciation guidance of mouth and tongue positions, and can effectively guide the learners to learn and practice English pronunciation.

“I Understood you, but there was this Pronunciation Thing…”: L2 Pronunciation Feedback in English/French Tandem Interactions

The role of corrective feedback (CF) in L2 development has been the topic of much discussion in SLA literature (see for example Sheen and Ellis 2011 for a recent overview). Researchers have focused their attention on CF provided either by language teachers or by fellow L2 learners, whereas relatively little is known about phonetic feedback offered in a non-institutional setting during peer-to-peer native/non-native interactions as is the case with tandem language learning. Tandem language exchanges represent a special learning environment, as each participant takes turns being the native and the non-native side of the dialogue. Thus, in contrast to the typical L2 learning setting, the hierarchical structure between the participants is fluid: the expert-novice power relationship evolves as the meeting progresses and the conversation switches from one language to the other. In order to see how the distinguishing characteristics of tandem learning (such as solidarity and reciprocity) shape the process of L2 phonetic development in their own specific ways, we collected an English-French Tandem Corpus as part of the SITAF project (Spécificités des Interactions verbales dans le cadre de Tandems linguistiques Anglais-Français), launched at the University of Sorbonne Nouvelle-Paris 3 in October 2012. We gathered linguistic data – both video and audio recorded – from face-to-face conversational exchanges held by 21 pairs of undergraduate students, with each ‘tandem’ consisting of a native speaker of English and a native speaker of French. The dialogues and reading passages were recorded on two occasions separated by a 3-month interval. The present paper offers a preliminary analysis of L2 pronunciation feedback on several renditions of the same text (The North Wind and the Sun), given to the French speakers by their English tandem partners. The passage was produced by each French participant three times: (1) during the ‘monitored’ reading, which was supervised by the English-speaking partner and which led to (2) the ‘second reading’ in the course of the first recording session, and then (3) the ‘final reading’ performed during the second recording session 3 months later. Data analysis will allow us to address the following questions relating to the study of phonetic corrective feedback:
 - What is corrected by the native-speaking partner (henceforth NS)? Segmental or prosodic errors? Phonemic or allophonic deviations?- What is the corrective strategy adopted by the NS? Is it explicit correction, recast, or elicitation?- What is the learner’s uptake after receiving feedback?We hope that our data brings a valuable and fairly unique contribution to SLA research, helping to establish which errors get corrected and how it may have implications for setting priorities in L2 pronunciation teaching.