Abstract:
The application of artificial intelligence (AI) has become pervasive across numerous domains in the current era of big data, and its contribution to the informatization of education is especially noteworthy. AI-assisted language education provides learners and teachers with diversified learning modes and resources. It not only enhances the personalised learning experience; but also fosters the development of independent learning skills while alleviating the burden on teachers. This paper mainly presents EAP Talk, a speech assessment system for spoken English based on AI and big data models, and concisely overviews its features and applicability in language education research.
Artificial Intelligence (AI) is increasingly employed in English teaching and learning, along with the continuous improvement and deep integration of speech recognition and synthesis technologies. Mobile applications based on these advancements present extensive opportunities and avenues for enhancing English education. This software enriches the interaction and practice in the process of language acquisition, through which students can carry out diverse activities such as interesting word memorisation, oral output practice, simulated listening exercises, writing correction, grammar revision, etc. The technological evolution of English learning software has progressed from conventional neural network models to expansive AI-driven architectures. Simultaneously, there has been a persistent focus on innovating and enhancing the quality of user interaction. Within this domain, EAP Talk, an AI English-speaking assessment system based on big data, speech recognition, natural language processing and deep learning, can automatically grade students’ speaking practice in real-time, thus providing a new path for improving speaking proficiency. Unlike other applications, EAP Talk focuses on learning and practicing English for academic purposes (EAP), deriving its content from academic English textbooks and authentic EAP classroom materials. Students gain access to a range of customizable functions tailored to their individual learning requirements, and key functions of EAP Talk include Reading aloud, Presentation, Discussion Expression, AI Chat, Word, etc. Through these different modes of practice, students can benefit from targeted improvements in speaking skills and presentation abilities based on individual needs, which in turn enhances their academic competitiveness.
Figure 1. Screenshot of EAP Talk Read Aloud Exercise
Each function of EAP Talk can provide diversified feedback based on students’ recordings. Take Reading aloud as an example, as shown in (Figure 1); students have the option to initially access a standard reading demonstration by clicking on “Example”. This demonstration guides word pronunciation, intonation, and serves as a model for imitation. Following this, students can begin their own recording by clicking “Start”. EAP Talk will automatically generate feedback scores (Figure 2) based on the recorded audio. The transcription of the audio into text is color-coded, representing various performance levels, e.g., yellow for perfect, green for good, and purple for inaccurate. Meanwhile, EAP Talk also provides feedback on the corresponding scores, including the dimensions of “fluency” (the fluency of English speech), “pronunciation” (the accuracy of word pronunciation, sentence enunciation, and pauses), and “Integrity” (completeness of the passages read aloud).
Figure 2. Screenshot of feedback of EAP Talk Reading aloud exercise
In the Presentation, students can respond to the avatar teacher’s questions and receive synchronous feedback. EAP Talk assigns the corresponding IELTS speaking scale scores and equivalent scores of other associated examinations. The transcribed text is then color-coded for clarity (as illustrated in Figure 3), with advanced vocabulary highlighted in yellow or green, while grammatical errors and pronunciation issues are marked in red. With this feedback mechanism, students can pinpoint their speaking strengths and weaknesses as well as formulate personalised learning strategies and examination preparation plans, thus enhancing the effectiveness of self-directed learning.
Figure 3. Screenshot of Feedback from EAP Talk Presentation Exercise
Academic Research Based on EAP Talk
The real-time and user-friendly nature of EAP Talk has generated substantial interest among English learners and educators, with a burgeoning application in academic research. English teachers can conduct both instruction and teaching-related practical research using EAP Talk. For example, Zou et al. (2023a) invited 70 college students to use several AI-supported English learning software, including EAP Talk, for English learning. They found that social network interactions can effectively improve learners’ oral expression in the AI-supported learning environment through a comparative study of the control and experimental groups. Survey and interview findings also indicated a predominant preference among students for employing such software to practice spoken English. In another empirical study conducted by Zou et al. (2023b), feedback mechanisms like score feedback and text color-coding of transcribed audio recordings, provided by EAP Talk, were highly favored by users as these feedbacks could visually help users comprehend their speaking proficiency. Furthermore, in a subsequent study on students’ acceptance of applying EAP Talk for speaking practice, a comprehensive analysis of questionnaires, interview data, and pre-and post-tests consistently demonstrated its efficiency in improving speaking performance. Additionally, the harmonious and real-time human-computer interaction environment facilitated by EAP Talk was found to be enjoyable (Zou et al., 2023c).
The aforementioned studies underscore the practical significance and academic value of EAP Talk in English education, revealing how AI contributes to the revolutionary development of personalised independent learning in intelligent English education. As AI technology continues to advance, the language learning process undergoes profound changes. The integration of AI not only diminishes language barriers in intercultural communication but also serves as a pivotal link that connects diverse backgrounds, fostering globalization through language as a unifying element.
References:
Zou, B., Guan, X., Shao, Y., & Chen, P. (2023a). Supporting Speaking Practice by Social Network-Based Interaction in Artificial Intelligence (AI)-Assisted Language Learning. Sustainability, 15(4), 2872. https://doi.org/10.3390/su15042872
Zou, B., Du, Y., Wang, Z., Chen, J., & Zhang, W. (2023b). An Investigation Into Artificial Intelligence Speech Evaluation Programs With Automatic Feedback for Developing EFL Learners’ Speaking Skills. SAGE Open, 13(3), 21582440231193818. https://doi.org/10.1177/21582440231193818
Zou, B., Lyu, Q., Han, Y., Li, Z., & Zhang, W. (2023c). Exploring students’ acceptance of an artificial intelligence speech evaluation program for EFL speaking practice: An application of the Integrated Model of Technology Acceptance. Computer Assisted Language Learning. https://doi.org/10.1080/09588221.2023.2278608
