Automated Assessment of Communicative Competence: An RNN- Driven AI Model for Student - Centered Intelligent Language Teaching

Abstract

Artificial Intelligence (AI) in language teaching has brought about revolutionary developments in student engagement, instructional design, and evaluation techniques. Processing Natural language (NLP)-based tests, gamified learning platforms, and automated tutoring systems are examples of AI-powered solutions that have greatly improved accessibility, efficiency, and personalization in education. These solutions use AI-driven algorithms to enable self-directed learning outside of typical classroom settings, provide immediate feedback, and dynamically modify learning content based on individual progress. This study aims to develop and evaluate an intelligent communicative language teaching model with artificial intelligence integration. The model incorporates three essential elements: Peer teaching, debate, group discussions, and role-playing that organizes classroom activities are all part of a five-stage CLT; An AI module that uses computerized translation, standard linguistic processing and recurrent neural networks; a speech-to-text conversion system that uses acoustical preprocessing, linguistic modeling, and acoustical modeling to transform spoken communication through textual that can be analyzed. The suggested intelligent communicative language teaching delivers a data-driven, scalable method to connect particular student-centered activities with quantifiable linguistic performance. The model's output is evaluated across two dimensions: English skills using standardized speaking for fluency accuracy coherence vocabulary and interaction; learning achievement measured by both pre- and post-test gain score.