An AI-Based Writing Assistant for IELTS : System Architecture and Implementation

• Published in: International Journal of Advanced Culture Technology(IJACT) Vol. 13; no. 2; pp. 179 - 186
• Main Authors: Linh Bui Duong Hoai, 이강희
• Format: Journal Article
• Language: English
• Published: 국제문화기술진흥원 30.06.2025
• Subjects: 과학기술학; Streamlit Application; IELTS Writing; Natural Language Processing; Computer-Assisted Language Learning
• ISSN: 2288-7202; 2288-7318

This paper presents the development of an AI-assisted system designed to support IELTS writing preparation by integrating large language models and retrieval-augmented techniques. The system was built using Python, with Streamlit providing the user interface, OpenAI's API enabling natural language processing capabilities, and FAISS combined with LangChain facilitating efficient semantic retrieval and conversational context handling. The tool allows learners to input IELTS writing tasks via direct text or uploaded PDF files, from which it extracts and processes text, tables, and images to generate personalized writing guidelines. It also enables automated grading aligned with IELTS writing criteria and offers improved essay suggestions targeting a specified band score. Integrating modular components and AI capabilities, the system offers a scalable framework for automated writing support in language learning contexts. This innovation advances the expanding domain of AI in educational settings and helps illustrate AI systems' potential to improve language skills.