🎬 IMDB Movie Reviews Sentiment Analysis

This project performs sentiment analysis on the IMDB Dataset of 50,000 movie reviews. The dataset is labeled with binary sentiments: positive or negative.

We go through a full machine learning pipeline using Natural Language Processing (NLP) techniques to classify review sentiment.

---

Dataset

• File: IMDB Dataset.csv
• Columns:
  • review: Text content of a user review
  • sentiment: Label (positive or negative)

---

Project Steps

1. Data Loading

   • Read and display the structure of the dataset.

2. Data Preprocessing

   • Remove HTML tags, punctuation, stopwords
   • Tokenization, lowercasing, and stemming

3. Exploratory Data Analysis

   • Sentiment distribution
   • Word clouds for positive and negative reviews
   • Review length analysis

4. Text Vectorization

   • Using TF-IDF for numerical feature extraction

5. Model Training

   • Trained a Logistic Regression model
   • Achieved accuracy over ~85% on test data

6. Evaluation

   • Classification report
   • Confusion matrix visualization

---

Visualizations

• Sentiment distribution bar plot
• Histogram of review lengths
• Confusion matrix heatmap

---

Requirements

pip install pandas numpy matplotlib seaborn scikit-learn nltk