Project Overview:

The “Customer Sentiment Analyzer for E-Commerce”** is a web-based system that collects, analyzes, and visualizes customer reviews from e-commerce platforms like Amazon or Flipkart. It uses Natural Language Processing (NLP) to determine whether reviews are positive, negative, or neutral and displays interactive sentiment insights on a dashboard.

This tool helps brands, sellers, and analysts understand customer opinions and product performance, assisting in decision-making and product improvement.

Key Features:

1. Review Scraper or CSV Upload

2. Sentiment Analysis using NLP (AI/ML)

3. Sentiment Dashboard with Pie/Bar Charts

4. Product-wise or Keyword-wise Sentiment Filters

5. Store Sentiment History in Database

6. Admin Panel to Manage Products & View Trends

Technology Stack:

Backend (Choose one):

• PHP (Laravel/Core)

• Node.js (Express + MongoDB)

• Java (Spring Boot + MySQL/PostgreSQL)

Frontend:

• HTML5 + CSS3

• Bootstrap (responsive design)

• JavaScript + AJAX

• Chart.js or D3.js for graphs

ML/NLP (Python):

• TextBlob / VADER for rule-based sentiment

• OR Scikit-learn + Naive Bayes / SVM

• Flask API to connect Python ML model with backend

Database:

• MySQL / PostgreSQL / MongoDB

Optional:

• Web scraping using BeautifulSoup or Scrapy

• REST API integration with e-commerce platforms

System Workflow:

1. Data Input:

   • Admin uploads review CSV file OR

   • Automatically fetch reviews using web scraper

2. Sentiment Processing:

   • Python ML model processes each review

   • Assigns labels: Positive / Negative / Neutral

3. Database Storage:

   • Stores raw review, product info, and sentiment

4. Dashboard Visualization:

   • Pie chart: % of each sentiment

   • Bar chart: Sentiment per product

   • Filters by rating, product, or date

Example ML Flow: