YouTube Comment Analysis – NLP Mini Project

This repository contains the dataset and Python scripts used to collect, clean, analyze, and visualize YouTube comments from the video:

"2025’s Most Important Career Podcast – Make Money Using AI | The Ranveer Show"
Video Link: https://youtu.be/YG1sW00jwLY

---

📂 Dataset

The dataset was extracted using the YouTube Data API and then cleaned using Python. After preprocessing, the final cleaned dataset file is:

• youtube_comments.csv
• youtube_comments_cleaned.csv (after text cleaning)

---

🛠️ Python Scripts Used

This project uses five Python files, each handling a different stage of the NLP workflow.

1. comments.py – YouTube Comment Scraper

• Uses YouTube Data API (googleapiclient)
• Fetches all top-level comments from the video
• Stores them into youtube_comments.csv

2. dataLoader.py – Dataset Loader

• Loads the CSV file
• Renames columns properly for consistent processing

3. cleaning.py – Text Preprocessing

• Converts text to lowercase
• Removes punctuation
• Tokenizes words
• Removes English stopwords
• Creates a new column cleaned in the DataFrame

4. analysis.py – NLP + Visualization

Contains:

• Word Frequency Calculation
• Bar Plot of Top 20 Words
• Sentiment Analysis using TextBlob
• Sentiment Distribution Graph

5. main.py – Project Pipeline

Runs the entire workflow:

1. Load dataset
2. Clean text
3. Show word frequencies
4. Plot frequency graph
5. Perform sentiment analysis
6. Plot sentiment graph

---

📁 Repository Structure