🎭 Sentiment Analysis on IMDb Reviews

This project applies Natural Language Processing (NLP) techniques to classify movie reviews from IMDb into positive or negative sentiment categories. It follows a two-phase approach:

• 🔍 Phase 1: Text preprocessing and exploratory analysis using R and tidyverse libraries.
• 🤖 Phase 2 (in progress): Sentiment classification models in Python using Scikit-learn and PyTorch.

📁 Dataset

• IMDb Reviews Dataset (Kaggle)
• 50,000 reviews labeled as positive or negative.

📌 Project Phases

🧪 Phase 1: Text Processing in R

• Text cleaning: lowercasing, stopword removal, HTML cleanup
• Tokenization and lemmatization
• POS tagging with udpipe
• Visualization: word clouds, bar charts, n-gram analysis
• 📈 Full EDA Notebook on RPubs

🧠 Phase 2: Modeling in Python (Coming Soon)

• Data exported as IMDB-cleaned.csv
• Model candidates:
  • Logistic Regression
  • Naive Bayes
  • Support Vector Machines
  • PyTorch-based classifier
• Metrics: Accuracy, F1-score, ROC-AUC

📦 Deliverables

• sentiment-analysis.Rmd: Full EDA notebook in R
• IMDB-cleaned.csv: Preprocessed dataset
• model_sentiment.py: Sentiment classification model (planned)
• Streamlit / Flask deployment (planned)

🚀 Deployment Ideas

• Build an interactive dashboard (Streamlit)
• Deploy a REST API using FastAPI
• (Optional) Real-time batch sentiment processing with Apache Spark

📚 Skills & Tools Used

• R: tidyverse, tidytext, udpipe, ggplot2, SnowballC
• Python: scikit-learn, NLTK, PyTorch (planned)
• EDA & Reporting: R Markdown, RPubs

🧠 Author

Manuel Alejandro Matías Astorga
Data Scientist | Physicist | Machine Learning Enthusiast
📄 Portfolio Website · LinkedIn

---

✨ Feel free to fork, contribute or reach out if you're working on similar projects!