Overview
Key Features
- User Authentication: Supports Google and ORCID-based sign-ins.
 - Personalized Recommendations: AI-driven suggestions based on user reading history.
 - Bookmarking System: Easily save and access important research papers.
 - Journal Subscriptions: Follow journals and receive real-time updates.
 - Advanced Search: Quickly find relevant papers and publications.
 - AI-Powered Summarization: Automatically generated paper summaries for quick insights.
 
Architecture
- Frontend: Flutter-based UI ensuring a smooth and responsive user experience.
 - Backend: Built with Node.js and Express, handling authentication, data storage, and recommendations.
 - Database: MongoDB for storing user preferences, research papers, and bookmarks.
 
Challenges and Learnings
- Efficient data storage: Optimizing MongoDB queries for fast retrieval.
 - Smooth user experience: Using Flutter's state management for seamless navigation.
 - Scalability: Ensuring the recommendation system performs well with large datasets.
 
Deep Dive: The Architecture of Kerosene
1️⃣ Application Structure
- Flutter Mobile App 📱 (Primary research tool)
 - Next.js Website 🌐 (Landing page + bookmark sharing)
 - Node.js Backend ⚡ (API + AI processing)
 - Population Script 🛠️ (Fills database with research papers)
 
2️⃣ Mobile App: Features & UI Flow
🏁 User Onboarding & Authentication
- Users sign in via Google or ORCID.
 - Onboarding screens (with interactive categories) personalize recommendations.
 
🔍 AI-Powered Research Paper Discovery
- Home Screen: "For You" tab (AI-based recommendations) & "Following" tab (journals user follows).
 - Search Screen: Keyword-based + AI-boosted search.
 - Paper Details Screen: Shows AI-generated summary & metadata.
 - Bookmarks & Journal Subscriptions: Papers can be saved for later reading, and journals can be followed.
 
⚙️ User Profile & Preferences
- Users can update reading preferences, manage bookmarks, and view followed journals.
 
3️⃣ Website: Bookmark Sharing & Legal Pages
- Acts as a landing page for Kerosene.
 - Allows users to share bookmarks using a Bookmark ID, making collaboration easier.
 
4️⃣ Backend API: Node.js + MongoDB
✔️ User Authentication: Login via Google or ORCID.
✔️ Recommendation System: AI-powered research paper suggestions.
✔️ Bookmarks & Journal Management: Save and retrieve saved content.
✔️ Search API: Supports keyword-based queries.
✔️ Journal Subscription System: Follow/unfollow journals.
✔️ AI-Powered Summarization: Extracts key insights from papers.
🔗 Key API Routes
- Authentication: Handled by ./routes/auth.
 - Papers: Handled by ./routes/papers.
 - Bookmarks: Handled by ./routes/bookmark.
 - Journal: Handled by ./routes/journal.
 
5️⃣ AI-Powered Enhancements
🔹 Paper Summarization – Generates concise abstracts.
🔹 Keyword Extraction – Improves search accuracy.
🔹 Personalized Recommendations – AI filters relevant research papers.
