Interactive web-based text-to-speech and speech recognition media for enhancing Arabic listening proficiency

Background: Arabic listening lessons often rely on passive audio repetition, leaving students with little opportunity to check their understanding or receive immediate feedback. This condition slows their progress, especially in recognizing key phonemes and responding accurately. Aims: This study aims to develop and validate a web-based interactive listening tool using text-to-speech (TTS) and speech-recognition (ASR) to support students’ Arabic listening proficiency. Methods: The research followed the Dick, Carey & Carey model through the stages of analysis, design, development, and expert validation. Data were collected through needs analysis, classroom observations, and interviews with teachers to identify learners’ challenges in Arabic listening. After the prototype was created, expert judgment sheets were used to gather validation data from design and content specialists. The responses were then analyzed using percentage-based validity scoring to determine the feasibility of the media. The prototype integrated web-based TTS and ASR modules, specifying platform architecture, speech engines, supported languages, and technical parameters to ensure replicability. Results: The results show that the media supports active listening practice and helps address common learner difficulties, particularly in phoneme recognition and comprehension. TTS provided clear and natural audio, while ASR offered direct corrective feedback. Expert assessments indicated high feasibility, with a design validity score of 92% and content validity of 87%. Implications: The findings indicate that TTS–ASR integration can ease teachers’ correction workload and help students practice more independently, offering practical potential for broader classroom use.