Examor is a web application designed to help students, scholars, and lifelong learners reinforce their knowledge through AI-powered self-testing. It allows users to upload notes and documents, from which it generates questions, assesses user answers using LLMs, and schedules reviews based on a simplified Ebbinghaus memory model.

How It Works

The core of Examor leverages LLMs (GPT-4 recommended) to generate questions from user-provided documents and to score user answers. It implements a simplified Ebbinghaus memory curve where higher scores lead to longer review intervals. Users can select question types and roles for varied assessment experiences. The system prioritizes knowledge retention by continuously prompting users with questions derived from their own learning materials.

Quick Start & Requirements

• Install via Docker: docker compose up
• Prerequisites: Docker installed, ports 51717, 51818, and 52020 available.
• Access: http://localhost:51818
• Documentation: Best Documentation Practices

Highlighted Details

• AI-powered question generation and answer assessment.
• Simplified Ebbinghaus memory model for spaced repetition.
• Role selection for diverse question generation and assessment.
• Question Bank feature for importing pre-existing questions from open-source documents.
• Notes Management for organizing uploaded documents and associated questions.

Maintenance & Community

The project is in an early stage with ongoing refactoring to Next.js (dev-next branch). Contributions are welcomed via issues and pull requests.

Licensing & Compatibility

• License: AGPL-3.0
• Compatibility: AGPL-3.0 is a strong copyleft license, requiring derivative works to also be open-sourced under the same license. Commercial use or linking with closed-source applications may require careful consideration of license obligations.

Limitations & Caveats

The project is in an early stage with ongoing refactoring, potentially leading to instability or breaking changes. Initial loading and certain page transitions may be slow due to module optimization. The Ebbinghaus implementation is simplified and planned for future optimization. The update process involves manual data export/import and container rebuilding.