Scholarlink

This AI-powered system automates the creation and evaluation of educational questions, supporting scalable and personalized learning. Using the Llama 3.3 70B model via Groq Cloud, it rapidly generates text, multiple-choice, and coding questions based on user-defined difficulty levels and targeted skills. Through Retrieval-Augmented Generation (RAG), it searches uploaded PDFs for relevant content. If unavailable, it relies on a built-in knowledge base to ensure coverage. Tools like LangGraph, LlamaIndex, Hugging Face embeddings, and ChromaDB manage workflows, structure documents, analyze semantics, and store vectors efficiently. During evaluation, it scores responses by comparing them with source content or the pretrained model, providing detailed feedback. This approach streamlines assessments, supports diverse subject matter, and reduces educator workload while identifying learner-specific gaps to enhance outcomes.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-Powered Question Generation and Evaluation Framework for Enhanced Educational Assessments

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Pavan Kumar Reddy Gaddam,
  • Samith Reddy Kandala,
  • G. Kavita

摘要

This AI-powered system automates the creation and evaluation of educational questions, supporting scalable and personalized learning. Using the Llama 3.3 70B model via Groq Cloud, it rapidly generates text, multiple-choice, and coding questions based on user-defined difficulty levels and targeted skills. Through Retrieval-Augmented Generation (RAG), it searches uploaded PDFs for relevant content. If unavailable, it relies on a built-in knowledge base to ensure coverage. Tools like LangGraph, LlamaIndex, Hugging Face embeddings, and ChromaDB manage workflows, structure documents, analyze semantics, and store vectors efficiently. During evaluation, it scores responses by comparing them with source content or the pretrained model, providing detailed feedback. This approach streamlines assessments, supports diverse subject matter, and reduces educator workload while identifying learner-specific gaps to enhance outcomes.