Twitter Sentiment Analysis Tool

Overview

This project is a real-time Twitter Sentiment Analysis tool built using Python, NLP, Streamlit, and Nitter. It allows users to analyze sentiments by either typing custom text or fetching live tweets from any Twitter profile—without relying on the Twitter API!

Features

• Scrape live tweets using ntscraper (Nitter-based scraping)
• Perform sentiment analysis using a custom-trained NLP model
• Interactive web-based interface using Streamlit
• Real-time visualization of sentiment trends

Tech Stack

• Python: Core programming language
• ntscraper: Scrapes tweets without the Twitter API
• NLTK/TextBlob/VADER: For sentiment analysis
• Streamlit: Web application framework for UI
• Matplotlib/Seaborn: Visualization of sentiment trends

Installation

Prerequisites

Ensure you have Python installed (>=3.7). Install dependencies using:

pip install ntscraper streamlit nltk textblob vaderSentiment matplotlib seaborn

Clone Repository

git clone https://github.com/yourusername/twitter-sentiment-analysis.git
cd twitter-sentiment-analysis

Usage

Run the Streamlit App

streamlit run app.py

Analyzing Tweets

1. Custom Input: Enter text and get sentiment classification (Positive/Neutral/Negative)
2. Fetch Live Tweets: Enter a Twitter username, and the app scrapes recent tweets to analyze sentiments

Project Structure

.
├── app.py               # **Streamlit app**
├── sentiment_model.py   # **Custom sentiment analysis model**
├── scraper.py           # **Nitter-based tweet scraper**
├── requirements.txt     # **Dependencies**
├── README.md            # **Project documentation**

Sentiment Analysis Methodology

• TextBlob: For polarity and subjectivity analysis
• VADER: Optimized for short text like tweets
• Custom Model: (Optional) Train a machine learning model using NLP techniques

Deployment

Deploy on Streamlit Cloud

1. Push the project to a GitHub repository
2. Go to Streamlit Cloud
3. Deploy the repository

Deploy on AWS (Optional)

Use AWS EC2 or Lambda to host the app for wider accessibility.

Future Improvements

• Improve sentiment classification with a fine-tuned ML model
• Add support for multilingual sentiment analysis
• Display time-series sentiment trends for a user’s tweets

Contributing

Pull requests are welcome! Please fork the repository and submit a PR.

License

This project is open-source and licensed under the MIT License.