Sentiment
Analysis
Understanding what people feel
A tool that analyzes emotional patterns in text, helping understand sentiment and tone. Built to make emotional intelligence more accessible through natural language processing.
Project Overview
This project involves analysing customer reviews from platforms such as Amazon and Flipkart to determine whether the sentiment expressed is positive, negative, or neutral.
I use natural language processing (NLP) techniques such as tokenisation, stopword removal, stemming, and TF-IDF vectorisation to preprocess the text data, and then train a Naive Bayes classifier to categorise the reviews and evaluate the model using a confusion matrix, accuracy, precision, and recall.
This project introduced me to text classification, NLP, and basic machine learning.
Implementation
- •Processed thousands of data points from e-commerce datasets
- •Achieved model accuracy of ~87%
- •Used confusion matrix, precision, and recall metrics for validation
- •Hands-on experience with pandas data handling and ML pipelines
Tech Stack
Timeline
Jul 2025 - Aug 2025