Glassmorphic Movie Explorer is a smart movie recommendation system that helps users discover relevant films based on genre preferences, keyword queries (e.g., "romance", "crime", "space"), and content similarity. Built with a focus on both intelligent retrieval and sleek user experience, the system mimics modern streaming platforms that personalize content using data-driven insights.
The core system blends traditional Information Retrieval (IR) methods with Machine Learning (ML) to provide ranked movie suggestions. The backend handles text preprocessing, BM25 scoring, and TF-IDF vectorization to compute initial relevance. Further, we enhance recommendation quality using ML models trained on user behavior data.
Users interact via a glassmorphic web interface that supports genre selection, keyword input, result customization, and sort-by filters (e.g., rating, year, ML relevance).
Key Features:
- Hybrid recommendation using IR + ML ranking
- Genre and keyword-based filtering
- Real-time score computation and relevance sorting
- Lightweight Flask backend with interactive HTML/CSS frontend
Technologies Used:
- Python, Flask, HTML/CSS/JS
- Libraries:
scikit-learn,xgboost,rank_bm25,nltk,pandas,TfidfVectorizer - Dataset: MovieLens 100K
- Metrics: nDCG, Precision@K, Recall
Authours:
- Arnav Sharma
- Katya Chadha
- Amisha Mittal