A multi-agent AI system using a BERT CEFR classifier in a generate-evaluate-regenerate loop produced more level-appropriate sentences (87.4% vs 54.1%), but its evaluation uses the same classifier that does the filtering, and a 3-person pilot found no statistically significant FLA reduction.
Exploring student perceptions of language learning affordances of Large Language Models: A Q methodology study,
1 Pith paper cite this work, alongside 7 external citations. Polarity classification is still indexing.
1
Pith paper citing it
7
external citations · OpenAlex
fields
cs.HC 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Towards Reducing Foreign Language Anxiety Using Level-Appropriate Embodied Conversational Agents
A multi-agent AI system using a BERT CEFR classifier in a generate-evaluate-regenerate loop produced more level-appropriate sentences (87.4% vs 54.1%), but its evaluation uses the same classifier that does the filtering, and a 3-person pilot found no statistically significant FLA reduction.