PaperHub is a scalable hybrid research paper recommender system that intelligently matches scholarly queries with relevant papers using a combination of semantic similarity (via Sentence-BERT) and citation-based scoring. It supports both category-based and cluster-based filtering using a trained KMeans model.
- 🔍 Query understanding using Sentence-BERT embeddings
- 🧠 Hybrid scoring using citation count + semantic similarity
- 📦 Optional KMeans-based cluster filtering
- 🖼️ Visually rich, responsive frontend built with Next.js + TailwindCSS
- 💾 Bookmark functionality using local storage
- 📄 PDF and source links for each recommended paper
To run the full stack locally, follow the instructions below.
git clone https://github.com/your-username/PaperHub.git📦 2. Install Dependencies 🔧 Backend (Python)
pip install -r requirements.txt🌐 3. Frontend (React + Next.js)
cd paperhub-ui
npm install🧠 4. Run the App
python app.py💻 5. Start the Frontend
cd paperhub-ui
npm run devNow visit 👉 http://localhost:3000 in your browser to explore the app.
