Sentiment Analysis Using Machine Learning

Sentiment Analysis Using Machine Learning and Deep Learning

I developed a text-classification project to compare traditional machine learning algorithms with a deep learning model for sentiment analysis.

The project covers text preprocessing, feature extraction, model training, performance evaluation, and saving trained models for reuse.

What I Built

Text preprocessing: Prepared review text for model training.

Feature extraction: Used TF-IDF for traditional machine learning models and tokenized sequences for the GRU model.

Model comparison: Evaluated Logistic Regression, Multinomial Naive Bayes, Linear SVM, Random Forest, and GRU.

Performance evaluation: Generated classification reports and visualizations to compare model results.

Model persistence: Saved trained models and preprocessing components for reuse.

Tech Stack

Python · Pandas · NumPy · Scikit-learn · TensorFlow/Keras · NLP · TF-IDF · Matplotlib · Joblib

What This Project Demonstrates

This project gave me practical experience building and evaluating NLP pipelines, comparing different modelling approaches, and examining how model choice affects text-classification performance.

Source Code