Finding the next book to read can be overwhelming. To solve this, I built BookVerse AI, a content-based book recommendation system web application with a modern glassmorphic dark-theme user interface.
Tech Stack
Frontend: Streamlit, custom Vanilla CSS, and Google Fonts (Outfit & Inter)
Machine Learning: Scikit-Learn (TF-IDF Vectorization and Cosine Similarity)
Data Handling: Pandas and NumPy
_*Visualizations: *_Plotly Express
Key Features
- AI Recommendation Engine: Computes similarity scores using Cosine Similarity on TF-IDF features of book descriptions, authors, and genres.
- Context-Aware Explanations: Generates a dynamic "Why We Recommend This" description showing matching themes, authors, and genres.
- Mood & Genre Discovery: An interactive mood selector (e.g., Happy, Motivated, Curious) maps user emotions to target book matches.
- Reading Dashboard: Analytics showing reading history, favorite genres, and mood statistics.
source code:
The project is fully open-source and available on GitHub:
https://github.com/pmarkhedkar-bit/bookverse-ai