Sentiment analysis is a natural language processing (NLP) technique used to determine whether data is positive, negative, or neutral.

Project Overview

The Sentiment-Analysis-NLP project focuses on building a sentiment analysis model for an e-commerce platform. The primary objective is to predict the binary rating (positive or negative) of product reviews based on their textual content. By leveraging machine learning techniques, the model classifies reviews and provides a submission-ready CSV file for evaluation.

Dataset Description

The project utilizes a dataset comprising product reviews, which includes the following features:

• Review_Title: The title of the product review.
• Review: The detailed text of the product review.
• Rating: The binary rating indicates positive or negative sentiment.

For modeling purposes, the Review_Title and Review are combined to form a single feature, providing a better context for sentiment analysis.

Methodology

The project employs a machine learning approach to classify the sentiment of product reviews. The key steps involved are:

1. Text Preprocessing: Combining Review_Title and Review into a single feature to capture the full context of the review.
2. Feature Extraction: Utilizing Term Frequency-Inverse Document Frequency (TF-IDF) vectorization to convert text data into numerical features suitable for modeling.
3. Modeling: Implementing a Logistic Regression classifier to predict the sentiment of the reviews.
4. Evaluation: Assessing the model's performance using metrics such as F1-Score and accuracy.

Results

The Logistic Regression model achieved the following performance metrics on the validation set:

• F1-Score: 0.987
• Accuracy: 98%

These results indicate a high level of precision and reliability in classifying the sentiment of product reviews.