🎬 MongoLens: Insights into Movie Trends with MongoDB & Python

This project explores the MovieLens dataset using MongoDB for non-relational data handling and Python for querying, analysis, and visualization. It aims to uncover patterns in movie popularity, user preferences, genre evolution, and rating behaviors over time.

📦 Project Structure

├── notebooks/
│ └── dataWrangling.ipynb # Data loading, merging, transformation & JSON export
│ └── extra.ipynb # MongoDB queries & visualizations using Python
├── merged_movies_ratings.json # Final processed dataset
├── dump/ # intermediate datasets
├── Pictures/
├── SMBUD Project - Federica Topazio.pdf # Full report
├── README.txt # Citations of dataset
└── README.md # This file

🗃 Dataset

MovieLens 100k dataset containing:

• ~100,000 ratings from 610 users
• 9,742 movies with genre metadata
• Timestamps for all ratings (from 1996 to 2018)

Main fields:

• userId, movieId, title, genres, rating, timestamp, release year

🔧 Technologies Used

• MongoDB: flexible document-based schema & aggregation queries
• Python (pandas, matplotlib, seaborn, pymongo): data wrangling and visualization
• Jupyter Notebooks: analysis and execution of Python code

🔍 Analysis Topics

✅ Data Wrangling (See: dataWrangling.ipynb)

• CSV loading, cleaning & transformation
• Genre parsing and timestamp formatting
• Normalization of ratings
• JSON export for MongoDB

📊 MongoDB Queries & Visualizations (See: extra.ipynb)

• Average rating per genre
• User behavior based on rating activity
• Most polarizing movies (variance in ratings)
• Rating trends over time (e.g., for Toy Story)
• Genre popularity and evolution over the years
• Temporal patterns (monthly ratings)
• Correlation between number of ratings and average rating

📈 Example Visualizations

• 🎭 Bar chart: average rating per genre
• 👤 Scatter plot: user engagement vs average rating
• 🧨 Variance bar chart: most polarizing movies
• 📆 Line plot: rating trends for specific movies over time
• 🧊 Heatmap: genre popularity changes by year
• 📉 Monthly rating trends

🧠 Key Learnings

• MongoDB is well-suited for semi-structured datasets with nested lists (e.g., genres)
• Aggregation pipelines enable complex, efficient querying
• Visual exploration of rating patterns can inform recommender systems

✍️ Author

Federica Topazio
Politecnico di Milano | Systems and Methods for Big and Unstructured Data (2023–2024)

📜 License

This project is licensed for academic and research purposes.