New RAG framework enhances Bengali MCQ generation and answer prediction

Researchers have developed a novel structure-aware retrieval-augmented generation (RAG) framework designed to improve the generation of academic multiple-choice questions (MCQs) and answer prediction in low-resource languages like Bengali. This framework models Bengali textbooks as hierarchical graphs and employs a graph neural network for passage retrieval, providing focused context to a large language model. Experiments show this approach surpasses standard dense retrieval methods in relevance and accuracy for both question generation and answer prediction. AI

IMPACT This research could lead to better educational tools and resources in low-resource languages, improving accessibility to academic content.

RANK_REASON The cluster contains an academic paper detailing a new methodology and experimental results. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• Abu Tarabin Surzo
• arXiv
• Bengali
• BengaliMCQ
• graph neural network
• Hugging Face
• large language model
• retrieval-augmented generation

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →