🤖 AI Review Classification & Management System

A beginner-friendly cloud computing mini-project that uses Natural Language Processing (NLP) to automatically analyze product reviews — detecting sentiment and product category — and provides an admin dashboard to manage review resolution status.

🌟 Project Overview

This system allows users to submit product reviews through a web interface. The backend automatically:

Detects sentiment (Positive / Negative / Neutral) using TextBlob NLP
Detects product category (Phone, Laptop, Chair, etc.) using keyword matching
Stores all reviews in a local SQLite database
Provides an Admin Dashboard where an operator can view, filter, search, and update the status of all reviews

This project is suitable as a cloud computing mini-project and can be deployed on AWS EC2 with minimal configuration.

✅ Features

User Review Page

Clean text input form for submitting reviews
Example review buttons for quick testing
Real-time character counter
Instant analysis result shown after submission
Flash messages for success/error feedback

Sentiment Analysis

Uses TextBlob polarity score
Three classes: Positive, Negative, Neutral

Product Category Detection

Keyword-based matching for 10 product categories
Fallback to "Other" if no category found

Admin Dashboard

View all submitted reviews in a table
Summary stats: total, positive, negative, unresolved counts
Filter by: Sentiment, Category, Status
Search by keyword in review text
Update status inline: Unresolved → Under Process → Resolved
Delete reviews
Auto-submit filters on dropdown change

🛠️ Technologies Used

Layer	Technology	Purpose
Backend	Flask 3.x	Web framework
Database	SQLite + SQLAlchemy ORM	Data storage
NLP	TextBlob	Sentiment analysis
ML/Data	scikit-learn, pandas	ML utilities (expandable)
NLP Extras	nltk, joblib	Text processing support
Frontend	Bootstrap 5	Responsive UI
Frontend	Vanilla CSS + JS	Custom styling & interactivity

📁 Project Structure

project/
│
├── app.py                  ← Main Flask application (all routes)
├── requirements.txt        ← Python dependencies
├── reviews.db              ← SQLite database (auto-created on first run)
├── README.md               ← This file
│
├── models/
│   ├── __init__.py
│   └── database.py         ← SQLAlchemy Review model + db init
│
├── utils/
│   ├── __init__.py
│   ├── sentiment.py        ← TextBlob sentiment analysis function
│   └── category.py         ← Keyword-based category detection function
│
├── templates/
│   ├── index.html          ← User review submission page
│   └── dashboard.html      ← Admin dashboard page
│
└── static/
    ├── css/
    │   └── style.css       ← Custom CSS (modern design)
    └── js/
        └── main.js         ← Client-side JS (form validation, UX)

🖥️ Installation & Setup (Windows / VS Code)

Prerequisites

Python 3.9+ installed (python.org)
VS Code installed (code.visualstudio.com)
Git (optional)

Step 1 — Open the Project in VS Code

File → Open Folder → Select your project folder

Step 2 — Open the Integrated Terminal in VS Code

Terminal → New Terminal   (or Ctrl + `)

Step 3 — Create a Virtual Environment

python -m venv venv

This creates an isolated Python environment inside a venv/ folder.

Step 4 — Activate the Virtual Environment

venv\Scripts\activate

You should see (venv) at the start of the terminal prompt.

Step 5 — Install Dependencies

pip install -r requirements.txt

Step 6 — Download TextBlob Language Data

TextBlob needs extra data for sentiment analysis:

python -m textblob.download_corpora

If that doesn't work, run: python -c "import nltk; nltk.download('punkt'); nltk.download('averaged_perceptron_tagger')"

▶️ How to Run the Project

python app.py

You should see:

[DB] Database initialized — tables created if not present.
 * Running on http://0.0.0.0:5000
 * Running on http://127.0.0.1:5000

Open your browser and go to:

User Review Page: http://127.0.0.1:5000
Admin Dashboard: http://127.0.0.1:5000/dashboard

To stop the server: press Ctrl + C in the terminal.

🗄️ How SQLite Works

SQLite is a lightweight, file-based database — perfect for development and small projects.

The database file reviews.db is automatically created the first time you run app.py
No separate database server is needed
The database stores all reviews in a table called reviews

Table columns:

Column	Type	Description
id	Integer	Auto-incremented unique ID
review_text	Text	The raw review entered by the user
sentiment	String	Positive / Negative / Neutral
category	String	Detected product category
status	String	Unresolved / Under Process / Resolved
created_at	DateTime	Timestamp when the review was submitted

SQLAlchemy ORM (Object Relational Mapper) allows us to work with the database using Python objects instead of raw SQL queries. For example:

# Saving a review
new_review = Review(review_text="Great phone!", sentiment="Positive", ...)
db.session.add(new_review)
db.session.commit()

# Fetching reviews
reviews = Review.query.filter_by(sentiment="Positive").all()

😊 How Sentiment Analysis Works

Library used: TextBlob

TextBlob is a Python NLP library. It analyzes text and returns a polarity score between -1.0 and +1.0:

Polarity Range	Meaning	Our Label
> 0.1	Positive	✅ Positive
< -0.1	Negative	❌ Negative
Between	Neutral	➖ Neutral

Code example (utils/sentiment.py):

from textblob import TextBlob

def analyze_sentiment(review_text):
    blob = TextBlob(review_text)
    polarity = blob.sentiment.polarity  # -1.0 to +1.0

    if polarity > 0.1:
        return "Positive"
    elif polarity < -0.1:
        return "Negative"
    else:
        return "Neutral"

Examples:

"The phone is amazing!" → polarity ≈ 0.8 → Positive
"The chair was broken and terrible" → polarity ≈ -0.7 → Negative
"The laptop battery is average" → polarity ≈ 0.0 → Neutral

🏷️ How Category Classification Works

Method: Simple keyword matching (no ML model needed)

A Python dictionary maps each category to a list of keywords:

CATEGORY_KEYWORDS = {
    "Phone":      ["phone", "smartphone", "mobile", "iphone"],
    "Laptop":     ["laptop", "notebook", "macbook"],
    "Chair":      ["chair", "seat", "stool"],
    # ... etc
}

The review is converted to lowercase, then scanned for each keyword. The first match wins.

Examples:

"My phone screen cracked" → finds "phone" → Phone
"The laptop is slow" → finds "laptop" → Laptop
"It arrived broken" → no keyword found → Other

This approach is simple, fast, and easy to extend by adding more keywords.

☁️ AWS EC2 Deployment Guide

Step 1 — Launch an EC2 Instance

Log in to AWS Console
Go to EC2 → Instances → Launch Instance
Choose Ubuntu Server 22.04 LTS (Free Tier)
Instance type: t2.micro (Free Tier)
Create or select a key pair (.pem file) — save it safely
Under Security Group, add a rule:
- Type: Custom TCP
- Port: 5000
- Source: 0.0.0.0/0 (allows public access)
Click Launch Instance

Step 2 — Connect to Your EC2 Instance (SSH)

On Windows, use PowerShell or Git Bash:

ssh -i "your-key.pem" ubuntu@<your-ec2-public-ip>

Replace your-key.pem with your key file path and <your-ec2-public-ip> with your EC2 public IP.

On first connect, type yes to accept the fingerprint.

Step 3 — Install Python on EC2

sudo apt update
sudo apt install python3 python3-pip python3-venv -y

Step 4 — Upload Your Project to EC2

Option A — Using SCP (from your Windows terminal):

scp -i "your-key.pem" -r "C:/path/to/project" ubuntu@<ec2-ip>:~/ai-review-system

Option B — Using Git:

git clone https://github.com/yourusername/ai-review-system.git
cd ai-review-system

Step 5 — Create Virtual Environment & Install Requirements

cd ~/ai-review-system
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
python -m textblob.download_corpora

Step 6 — Run the Flask App

python app.py

Step 7 — Access the App in Your Browser

Open your browser and go to:

http://<your-ec2-public-ip>:5000

Make sure port 5000 is open in your EC2 Security Group (Step 1).

Step 8 (Optional) — Keep the App Running in Background

nohup python app.py > output.log 2>&1 &

This keeps the app running even after you close the SSH session.

Step 9 (Optional) — Use a Process Manager (PM2 or Gunicorn)

For production, run with Gunicorn:

pip install gunicorn
gunicorn -w 4 -b 0.0.0.0:5000 app:app

📸 Screenshots

(Add screenshots here after running the application)

Page	Description
Review Page	User submits a review, sees result
Analysis Result	Sentiment and category displayed
Admin Dashboard	All reviews with stats and filter bar
Status Update	Operator changes review status inline

🚀 Future Improvements

User Authentication — Add login for admin dashboard
Advanced ML Model — Replace keyword matching with a trained TF-IDF or BERT classifier
Email Notifications — Notify admins when new negative reviews arrive
Charts & Analytics — Add Plotly/Chart.js charts on the dashboard
Pagination — Handle large numbers of reviews efficiently
Export to CSV — Let admin download review data
Multi-language Support — Detect sentiment in languages other than English
REST API — Expose endpoints for mobile/other frontends
PostgreSQL — Upgrade from SQLite to PostgreSQL for production
Docker — Containerize for easier deployment

📄 License

This project is for educational purposes. Free to use and modify.

Built with ❤️ using Flask · TextBlob · SQLAlchemy · Bootstrap 5

Name		Name	Last commit message	Last commit date
models		models
static		static
templates		templates
utils		utils
README.md		README.md
app.py		app.py
requirements.txt		requirements.txt
reviews.db		reviews.db

Latest commit

History

Folders and files

Repository files navigation