Submission for Hack The Ridge 2025.

A smart book recommendation engine powered by vector embeddings.

Discover books tailored to your interests! NovelNudge uses SentenceTransformers to embed book titles, descriptions, authors, and genres into vectors. It then finds the most similar books using cosine similarity and fuzzy title matching, delivering highly relevant recommendations in real-time.

• Search by book title or description
• Combines multiple factors (title, description, author, genre) into a weighted vector for better recommendations
• RapidFuzz integration for fuzzy matching of book titles
• Responsive web interface with live search results and thumbnails
• Easily extendable dataset and vector embeddings

• Python 3.10 — 3.12 recommended
• pip installed

git clone https://github.com/jmagali/NovelNudge.git

C:\Users\User\NovelNudge

pip install -r requirements.txt # This installs the required libraries

python app.py

Navigate to http://127.0.0.1:5000/ to use NovelNudge.

• Vectorization: Titles, descriptions, authors, and genres are encoded into vectors using SentenceTransformer.
• Combining Vectors: Each book's vectors are weighted and summed to form a combined vector.
• Search & Matching:
  • Title mode: RapidFuzz fuzzy matching selects a close match first.
  • Description mode: Input text is embedded and compared with all combined vectors.
• Recommendation: Top-N similar books are returned based on cosine similarity.
• Frontend Display: Results are displayed with thumbnails, authors, ratings, and similarity scores.

.
├── data/
│   ├── vectors/
│   │   ├── combined_vectors.npy
│   │   ├── vectorsAuthor.npy
│   │   ├── vectorsDes.npy
│   │   ├── vectorsGenre.npy
│   │   └── vectorsTitle.npy
│   ├── genre_list.csv
│   ├── processed_data.csv
│   └── unprocessed_data.csv
├── static/
│   ├── css/
│   │   └── style.css
│   ├── js/
│   │   └── app.js
│   └── img/
│       ├── book-education-learn-svgrepo-com.svg
│       └── github-mark.svg
├── templates/
│   └── index.html
├── .gitignore
├── .gitattributes
├── app.py
├── process_data.py
├── recommendations.py
└── requirements.txt
└── README.md

Dataset from Kaggle