WordPlay: An Agent Framework for Language Learning Games

ACL ARR 2024 April Submission410 Authors

Abstract: We introduce a novel framework, WordPlay, for building language learning games tailored to a learner's proficiency level. WordPlay combines playful mini-puzzle games with large language models and text-to-image models to address the challenge of balancing engagement and effective language practice. We showcase the framework's adaptability by implementing a wide variety of language learning games with diverse learning objectives. We evaluate WordPlay's ability to target different proficiency level by conducting experimental sessions with English language learners. A fine-tuned BERT-based model rates the difficulty of both LLM-generated and user responses according to Common European Framework of Reference (CEFR) learning levels. Our results demonstrate that WordPlay successfully elicits learner output aligned with targeted proficiency levels.

Paper Type: Short

Research Area: NLP Applications

Research Area Keywords: Language Modeling, NLP Applications

Contribution Types: NLP engineering experiment, Data analysis

Languages Studied: English

Submission Number: 410