Sentilytics

Multilingual YouTube Comment Sentiment Analysis Platform

A real-time sentiment analysis dashboard that extracts YouTube comments and classifies them using BERT and VADER — with AI-generated audience summaries and interactive analytics.

---

Overview

Sentilytics is a sentiment analysis platform that analyzes YouTube comments in real time. It fetches comments from any YouTube video, runs them through a BERT and VADER NLP pipeline, and visualizes audience sentiment through an interactive Streamlit dashboard.

The platform supports multilingual comments and generates AI-powered summaries of overall audience reaction, making it useful for content creators, marketers, and researchers.

---

Key Features

Sentiment Analysis Engine

• BERT-based sentiment classification (Positive / Neutral / Negative)
• VADER scoring for fast lexical sentiment estimation
• Multilingual comment support
• Confidence scores per classification

YouTube Comment Scraper

• YouTube Data API integration
• Bulk comment extraction from any public video
• Automated text preprocessing pipeline
• Language detection support

Interactive Dashboard

• Real-time sentiment breakdown charts (pie + bar)
• Searchable and filterable comment table
• Sentiment distribution by comment volume
• CSV export for further analysis
• Dark-themed modern UI built with Streamlit

AI Summary

• OpenAI-generated audience reaction summary
• Overall tone interpretation
• Key sentiment drivers highlighted

---

Screenshots

Dashboard

Sentiment Charts

Comments Analysis

---

System Architecture

YouTube Video URL
        |
YouTube Data API
        |
Comment Extraction (sentilytics_scraper)
        |
+----------------------------------+
|  Text Preprocessing              |
|  BERT Sentiment Classification   |
|  VADER Sentiment Scoring         |
|  OpenAI Summary Generation       |
+----------------------------------+
        |
Streamlit Analytics Dashboard
        |
CSV Export

---

Tech Stack

Category	Technologies
AI / NLP	BERT, Transformers, VADER
AI Summary	OpenAI API
Dashboard	Streamlit
Data Processing	Pandas, NumPy
Visualization	Plotly, Matplotlib
APIs	YouTube Data API v3
Language	Python
Tools	GitHub, VS Code

---

Getting Started

Prerequisites

• Python 3.9+
• YouTube Data API key (free at console.cloud.google.com)
• OpenAI API key (optional — for AI summary feature)

1. Clone the Repository

git clone https://github.com/AbiramiMuthiah/sentilytics.git
cd sentilytics

2. Install Dependencies

pip install -r requirements.txt

3. Set Environment Variables

# Windows
set YOUTUBE_API_KEY=your_youtube_api_key
set OPENAI_API_KEY=your_openai_api_key

# Mac/Linux
export YOUTUBE_API_KEY=your_youtube_api_key
export OPENAI_API_KEY=your_openai_api_key

4. Run the Dashboard

streamlit run live_youtube_analysis.py
# Opens at http://localhost:8501

---

Project Structure

sentilytics/
├── sentilytics/
│   └── sentilytics_scraper/   # YouTube comment extraction
├── assets/                    # Screenshots
│   ├── Dashboard.png
│   ├── sentiment-charts.png
│   └── comments-analysis.png
├── live_youtube_analysis.py   # Main Streamlit app
├── requirements.txt
└── README.md

---

Future Improvements

• Multi-platform support (Twitter/X, Reddit, Instagram)
• Real-time live stream comment analysis
• Emotion detection beyond positive/negative/neutral
• Fine-tuned multilingual BERT model
• User authentication and saved analysis history
• Cloud deployment on Streamlit Cloud or AWS

---

Author

Abirami Muthiah
Applied AI Engineer | NLP | Data Science

---

License

Licensed under the MIT License.