🎬 Movie Review Sentiment Analysis Platform

Educational & developer-friendly Flask + NLP app that aggregates movie reviews from Letterboxd, Rotten Tomatoes, OMDb, and YouTube trailer comments to perform sentiment analysis, keyword extraction, genre visualization, and interactive Plotly dashboards.

📌 Overview

This project was originally built for a college assignment and has been upgraded to a structured, extensible, and production-style educational repository. It demonstrates:

• Web scraping (ethical & rate-conscious patterns) for movie reviews
• Sentiment analysis using TextBlob, NLTK, optional profanity filtering, and basic keyword extraction
• Visualization with Plotly (distribution, polarity, keywords, genre breakdown)
• Comparison of audience mood before release (YouTube trailer comments) vs after release (user reviews)
• Clean project structure, Docker build, CI pipeline, tests, and contribution standards

⚠️ Disclaimer: This repository is for learning & research. Always review platform Terms of Service before scraping and avoid abusive traffic patterns.

🚀 Features

• 🔍 Enter a movie title – automatic metadata & poster retrieval via OMDb
• 🗂 Aggregated reviews (Letterboxd + Rotten Tomatoes users)
• ▶️ Pre-release sentiment via YouTube trailer comments
• 📊 Interactive charts: pie, bar, polarity histogram, genre distribution
• 🧠 Keyword extraction for positive & negative reviews
• ☁️ Word cloud generation
• 🧼 Profanity filtering & basic NLP cleanup (spaCy + NLTK)
• 🐳 Dockerized for reproducibility
• ✅ CI: linting, tests, security scanning

📊 Sample Screens

🧱 Architecture

Component	Purpose
app.py	Flask web server, routing & orchestration
imdb_scraper.py	Fetches metadata (OMDb) + Letterboxd + Rotten Tomatoes + similar movies
youtube_scraper.py	Trailer search + comment harvesting (YouTube Data API)
sentiment_analysis.py	Sentiment classification, keyword extraction, word cloud, plots
templates/	Jinja2 HTML templates for UI
static/	CSS & generated plot assets
config.py	Central environment & app configuration
tests/	Pytest-based unit tests

🛠 Tech Stack

• Python 3.11
• Flask, Plotly, TextBlob, NLTK, spaCy, WordCloud
• BeautifulSoup4, Requests
• Google API Client (YouTube Data API)
• Docker, GitHub Actions, Ruff, Black, Pytest

📦 Quick Start

1. Clone & Prepare

git clone https://github.com/jambhaleAnuj/senti_analysis_1.git
cd senti_analysis_1
python -m venv venv
venv\Scripts\activate  # Windows
pip install -r requirements.txt
pip install -r requirements-dev.txt  # optional for contributors

2. Configure Environment

Copy .env.example → .env and fill in:

OMDB_API_KEY=your_key
YOUTUBE_API=your_key
GEMMA_API_KEY = your_key

3. Run

python app.py

Open: http://127.0.0.1:5000

4. Docker (Optional)

docker build -t movie-sentiment .
docker run --rm -p 5000:5000 --env-file .env movie-sentiment

🧪 Testing

pytest -q

🧹 Code Quality

ruff check .
black .

🌐 SEO & Discoverability

Relevant keywords included: Movie Sentiment Analysis, Flask NLP, YouTube Comments Sentiment, Rotten Tomatoes reviews, Letterboxd scraping, Plotly dashboard, TextBlob sentiment, Python movie analytics, genre visualization, pre-release vs post-release audience mood.

🔐 Environment Variables

Variable	Purpose	Required
OMDB_API_KEY	Movie metadata & posters	Yes (fallback demo key used if absent)
YOUTUBE_API	Trailer comments sentiment	Optional (feature disabled if missing)
GEMMA_API_KEY	Trailer comments sentiment	Optional (feature disabled if missing)
FLASK_DEBUG	Enable debug mode	No

🧠 Sentiment & Keywords

Sentiment (TextBlob polarity): >0 positive, =0 neutral, <0 negative. Keywords: filtered tokens (stopwords removed, alphanumeric, length>2) frequency top 10 per class. Limitations: sarcasm, slang, named entities may skew results. Consider VADER / transformers for production.

🧭 Roadmap

• Caching (Redis) for repeated titles
• Transformer-based sentiment option
• Better keyword extraction (bigrams, POS filtering)
• PDF export of analysis
• Public JSON API endpoints

🤝 Contributing

See CONTRIBUTING.md. PRs welcome.

🛡 Security

No real keys committed. See SECURITY.md.