Movie Reviews Sentiment Analysis

Overview

This project implements sentiment analysis on movie reviews using both traditional Machine Learning and BERT-based approaches. The system can classify movie reviews as either positive (1) or negative (0) with high accuracy.

Project Structure

├── BERT accuracy/          # BERT model performance visualizations
├── ML models accuracy/     # ML models performance visualizations
├── NLP_Data/               # Dataset files
│   └── all_reviews.csv     # Combined dataset
├── NLP project.py          # Main project implementation
├── NLP project.ipynb       # Main project implementation as Jupyter Notebook
├── Documentation.pdf       # Detailed documentation
└── requirements.txt        # Project dependencies

Setup and Installation

1. Clone the repository:

git clone https://github.com/Mohammed2372/Movie-Reviews-Sentiment-Analysis.git
cd Movie-Reviews-Sentiment-Analysis

2. Create a virtual environment (recommended):

python -m venv venv
.\venv\Scripts\Activate

3. Install dependencies:

pip install -r requirements.txt

4. Download NLTK resources: The script will automatically download required NLTK resources on first run, or you can manually download them:

import nltk
nltk.download(['punkt', 'wordnet', 'stopwords', 'averaged_perceptron_tagger'])

Model Training

Run the main script to train both ML models and BERT:

python "NLP project.py"

Models

Traditional Machine Learning

• Logistic Regression
• Linear SVC
• Random Forest All models use TF-IDF vectorization with unigrams and bigrams.

BERT Model

• Base: textattack/bert-base-uncased-SST-2
• Fine-tuned for sentiment analysis
• Includes early stopping and model checkpointing

Performance

Machine Learning Models

• Results available in ML models accuracy/
• Classification reports for each model
• Comparative performance analysis

BERT Model

• Results available in BERT accuracy/
• Training loss curves
• Evaluation metrics
• Final test results

Documentation

For detailed information about:

• Data preprocessing steps
• Model architectures
• Training configurations
• Performance metrics
• Implementation details

Please refer to Documentation.pdf for more details.

Usage

To use the trained model for predictions (after training and saving the model):

from transformers import BertForSequenceClassification, BertTokenizer

# Load the model
model_path = "./bert_model"
model = BertForSequenceClassification.from_pretrained(model_path)
tokenizer = BertTokenizer.from_pretrained(model_path)

# Prepare text
text = "Your movie review here"
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=128)

# Get prediction
outputs = model(**inputs)
prediction = torch.argmax(outputs.logits, dim=-1)
sentiment = "positive" if prediction == 1 else "negative"