DeepTutor and PersonalExam
DeepTutor, an AI-powered personalized learning assistant, and PersonalExam, an AI-powered personalized exam system, are complementary tools within the educational learning RAG ecosystem, as DeepTutor likely focuses on delivering tailored learning content while PersonalExam specializes in assessing that learning through adaptive question generation, making them a natural pair for a comprehensive personalized educational experience.
About DeepTutor
HKUDS/DeepTutor
"DeepTutor: AI-Powered Personalized Learning Assistant"
Implements a dual-loop multi-agent architecture combining RAG, web search, and code execution for problem-solving with cited sources. Built on FastAPI (backend) and Next.js/React (frontend), it supports pluggable LLM and embedding providers with a unified prompt management system. Core modules handle document ingestion with Docling support, interactive visualization with session-based tracking, exam-style question generation, and cross-domain knowledge synthesis through deep research workflows.
About PersonalExam
sribdcn/PersonalExam
🧠 AI-powered Personalized Exam System: Integrating OpenPangu LLM, Knowledge Graph RAG, and BKT algorithm for adaptive question generation and recommendation. 基于LLM与知识图谱的智能个性化出题系统。
Implements dynamic knowledge graph construction via LLM-driven entity-relation extraction and automatic update detection through file hashing, while BKT algorithm tracks fine-grained mastery across hierarchical knowledge points to drive adaptive recommendation. Combines OpenPangu 7B for answer evaluation and semantic analysis with BGE embeddings and RAG-based retrieval to surface contextually relevant questions, all accelerated on Ascend NPU hardware.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work