IMDB 50K Movie Reviews Sentiment Classification

A comprehensive sentiment analysis project that classifies IMDB movie reviews as positive or negative using machine learning techniques. This project compares the performance of XGBoost against traditional linear classifiers on high-dimensional TF-IDF features.

🎯 Project Overview

This project implements binary sentiment classification on the IMDB 50K Movie Reviews dataset, focusing on:

• Exploratory Data Analysis (EDA) of review patterns
• Text preprocessing and feature extraction using TF-IDF
• Model comparison between XGBoost, Logistic Regression, and Linear SVC
• Performance evaluation and feature importance analysis

📊 Dataset

• Source: IMDB 50K Movie Reviews (Kaggle)
• Size: 50,000 reviews (25K positive, 25K negative)
• Features: Raw text reviews with binary sentiment labels
• Balance: Perfectly balanced dataset

🔧 Methodology

1. Data Preprocessing

• HTML tag removal
• Punctuation and noise cleaning
• Stopword removal
• Lemmatization
• Text normalization

2. Feature Engineering

• TF-IDF vectorization with unigrams and bigrams
• Maximum features: 100,000
• Sparse matrix representation

3. Model Implementation

• XGBoost: Primary gradient boosting classifier
• Logistic Regression: Linear baseline model
• Linear SVC: Support vector classifier baseline

📈 Results

Model	Accuracy	Precision	Recall	F1-Score	AUC-ROC
Linear SVC	90.87%	90.50%	91.49%	90.99%	96.88%
Logistic Regression	89.97%	88.96%	91.45%	90.18%	96.54%
XGBoost	86.18%	85.16%	87.89%	86.50%	93.96%

🔍 Key Findings

• Linear models outperform XGBoost on high-dimensional TF-IDF features
• Linear SVC achieved the highest accuracy at 90.87%
• All models demonstrate excellent discriminative ability (AUC-ROC ≥ 0.94)
• XGBoost shows strong recall but lower precision compared to linear models
• Linear decision boundaries prove more effective for this text classification task

🛠️ Technologies Used

• Python 3.x
• Machine Learning: XGBoost, Scikit-learn
• Data Processing: Pandas, NumPy
• NLP: NLTK
• Visualization: Matplotlib, Seaborn
• Text Processing: TF-IDF Vectorization

🛠️ Technologies Used

• Python 3.x
• Machine Learning: XGBoost, Scikit-learn
• Data Processing: Pandas, NumPy
• NLP: NLTK
• Visualization: Matplotlib, Seaborn
• Text Processing: TF-IDF Vectorization