Shranjeno v:
| Glavni avtor: | |
|---|---|
| Format: | Recurso digital |
| Jezik: | angleščina |
| Izdano: |
Zenodo
2026
|
| Teme: | |
| Online dostop: | https://doi.org/10.5281/zenodo.19709769 |
| Oznake: |
Označite
Brez oznak, prvi označite!
|
Kazalo:
- <p>In the present digital era, rapid advancements in technology have significantly transformed<br>the way people learn and communicate. The widespread availability of the internet and<br>smart devices has enabled the development of intelligent systems that provide personalized<br>learning experiences. Among various skills, effective communication in English has become<br>essential for academic success, career growth, and global interaction. However, many<br>learners face difficulties in achieving accurate pronunciation, fluency, and grammatical<br>correctness due to the lack of real-time guidance and individualized feedback.<br>Traditional language learning methods often rely on classroom-based teaching or static<br>digital content, which do not adequately address the unique learning pace and speaking<br>abilities of individual users. These limitations highlight the need for an intelligent and<br>interactive system that can assist users in improving their pronunciation skills through<br>continuous practice and evaluation.<br>The SpeakUp – AI Pronunciation Developer system is designed to overcome these challenges<br>by leveraging modern web technologies and artificial intelligence techniques. The<br>system integrates speech recognition, natural language processing, and audio analysis to<br>create a dynamic and user-centric learning platform. It enables users to practice pronunciation<br>by responding to system-generated questions, which are categorized into multiple<br>difficulty levels to ensure gradual skill enhancement.<br>The application captures user speech input through audio recordings and processes it<br>using a speech-to-text engine to obtain the transcribed content. The system then evaluates<br>the pronunciation using a hybrid scoring mechanism that combines textual similarity and<br>phoneme-level analysis, ensuring a more accurate assessment of speech quality. Additionally,<br>the system provides real-time grammar correction, detailed feedback, and generates<br>reference audio for correct pronunciation using text-to-speech technology.</p> <p>The platform follows a role-based architecture consisting of administrators, tutors, and<br>students. Administrators manage the system and approve tutors, tutors guide students<br>by generating and monitoring learning content, and students interact with the platform to<br>practice and improve their pronunciation skills. The system also includes advanced features<br>such as performance analytics, achievement tracking, and interactive communication<br>modules between tutors and students.<br>By integrating artificial intelligence with an intuitive web interface, the proposed system<br>offers a scalable, cost-effective, and accessible solution for pronunciation learning. This<br>approach not only enhances user engagement but also provides a structured and efficient<br>way to improve speaking skills, thereby addressing a critical need in modern language<br>education.</p>