Saved in:
| Main Authors: | , |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2025
|
| Online Access: | https://doi.org/10.5281/zenodo.17200797 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <div dir="auto"> </div> <div dir="auto"> <h2>Overview</h2> </div> <p dir="auto"><strong>A</strong>n open-source, web-based application designed to fully automate the <strong>primary study selection process</strong> in <strong>Systematic Literature Reviews (SLRs)</strong>. Built with <strong>Angular</strong> for the frontend and <strong>Node.js</strong> for the backend, This applcation leverages the power of <strong>GPT-4</strong> to dynamically retrieve, filter, and analyze research papers across multiple disciplines.</p> <p dir="auto">This tool aims to:</p> <ul> <li>Eliminate manual efforts in SLRs.</li> <li>Provide transparent, user-controlled automation.</li> <li>Enhance cross-domain applicability.</li> </ul> <div dir="auto"> <h2>Key Features</h2> </div> <ul> <li><strong>Dynamic Paper Retrieval</strong>: Automates the search and retrieval of research papers using customizable search strings.</li> <li><strong>Full Automation</strong>: From data collection to study selection—no manual uploads required.</li> <li><strong>Customizable Criteria</strong>: Flexible inclusion/exclusion parameters tailored to specific research needs.</li> <li><strong>Integrated Chat</strong>: Interact with an LLM trained on your selected studies for deeper insights.</li> <li><strong>SLR History & Traceability</strong>: View, edit, and manage all past SLRs for continuous improvement.</li> <li><strong>Cross-Disciplinary Support</strong>: Tested in medical sciences, engineering, and social sciences.</li> </ul> <div dir="auto"> <h2>Performance Highlights</h2> </div> <ul> <li><strong>Precision Rate</strong>: 82.6% (High accuracy in identifying relevant studies)</li> <li><strong>Retention Rate</strong>: 2.9% (Selective and rigorous screening process)</li> <li><strong>Sensitivity (recall) </strong>: 92.7</li> </ul> <p dir="auto">These metrics emphasize WebLit’s focus on <strong>quality over quantity</strong> in study selection.</p> <div dir="auto"> <h2>Tech Stack</h2> </div> <ul> <li><strong>Frontend:</strong> <a href="https://angular.io/" rel="nofollow">Angular</a></li> <li><strong>Backend:</strong> <a href="https://nodejs.org/" rel="nofollow">Node.js</a></li> <li><strong>LLM:</strong> GPT-4</li> </ul> <div dir="auto"> <h2>Getting Started</h2> </div> <div dir="auto"> <h3>Prerequisites</h3> </div> <ul> <li><strong>Node.js</strong> (v14 or above)</li> <li><strong>Angular CLI</strong> (v15 or above)</li> <li><strong>npm</strong> (v6 or above)</li> </ul> <div dir="auto"> <h3>Installation</h3> </div> <ol> <li> <p dir="auto"><strong>Environment Setup</strong> Rename .env-example to .env Add the required environment variables in .env (Ensure you have API keys or credentials if needed)</p> </li> <li> <p dir="auto"><strong>Backend Setup</strong></p> <div dir="auto"> <pre>npm install && npm run dev </pre> <div> </div> </div> </li> <li> <p dir="auto"><strong>Frontend Setup</strong></p> <div dir="auto"> <pre>cd client-v2 npm install ng serve --open</pre> </div> </li> </ol>