MetaScreener  by ChaokunHong

AI-powered literature screening tool

created 2 months ago
962 stars

Top 39.1% on sourcepulse

GitHubView on GitHub
Project Summary

MetaScreener is an AI-powered tool designed to significantly accelerate literature screening for systematic reviews and other research projects. It targets researchers and institutions aiming to reduce the time and effort associated with manually reviewing large volumes of academic papers, promising to cut screening time by up to 89%.

How It Works

MetaScreener employs a human-AI partnership model, allowing users to leverage various large language models (OpenAI, Claude, Gemini, DeepSeek) for abstract and full-text analysis. Users define screening criteria based on established research frameworks (e.g., PICOT, SPIDER) and upload documents (RIS, PDFs). The system then processes these, providing reasoned decisions and extracting data, with a strong emphasis on privacy by processing data locally.

Quick Start & Requirements

  • Online: metascreener.net (Recommended, no installation)
  • Local: Requires Python 3.10+. Installation involves git clone, pip install -r requirements.txt, and python app.py. OCR for PDFs requires Tesseract.
  • Setup Time: ~17 minutes (including API key setup).
  • Cost: $0.07-$1.50 per review (API usage).

Highlighted Details

  • Achieves 95-97% sensitivity and 85-92% specificity, with Cohen's Kappa of 0.85+.
  • Supports multiple LLM providers (OpenAI, Claude, Gemini, DeepSeek) for flexibility.
  • Integrates automated quality assessment tools (AMSTAR 2, Cochrane RoB 2, QUADAS-2).
  • Offers four core functionalities: Abstract Screening, Full-Text Analysis, Data Extraction, and Quality Assessment.

Maintenance & Community

Developed by researchers from the University of Oxford and the Oxford University Clinical Research Unit. Active community engagement via GitHub Issues and Discussions. Roadmap includes active learning, a REST API, and multi-user collaboration.

Licensing & Compatibility

The project is marked as "Open Source" but does not specify a license type in the README. This requires clarification for commercial use or linking with closed-source projects.

Limitations & Caveats

The specific open-source license is not clearly stated, which could impact commercial adoption. While it supports multiple LLMs, performance and cost will vary significantly between providers. The tool is primarily focused on English-language literature, with multilingual support planned for the future.

Health Check
Last commit

2 months ago

Responsiveness

Inactive

Pull Requests (30d)
0
Issues (30d)
0
Star History
982 stars in the last 90 days

Explore Similar Projects

Feedback? Help us improve.