This project is a real-time sentiment analysis pipeline for YouTube comments. It collects comments from YouTube videos, performs sentiment analysis using the Gemini API, and shares negative results on a Telegram channel.

• Introduction
• Features
• Getting Started
  • Prerequisites
  • Installation
  • Configuration
  • Running the Project
• Project Structure
• License

This project aims to create a real-time pipeline that collects comments from YouTube videos, performs sentiment analysis on these comments using the Gemini API, and shares the negative results on a Telegram channel. The pipeline consists of three main components:

1. Comment collection and transfer to Kafka.
2. Sentiment analysis using the Gemini API.
3. Telegram bot for sharing negative comments.

• Real-time collection of YouTube comments.
• Sentiment analysis using the Gemini API to classify comments as positive, negative, or neutral.
• Automatic sharing of negative comments on a specified Telegram channel.

• Python 3.6 or higher
• Kafka
• KSQL
• YouTube Data API key
• Gemini API key
• Telegram Bot API key

1. Clone the repository:

   git clone https://github.com/bunyaminonum/youtube-comments-sentiment-analyses.git
   cd youtube-comments-sentiment-analyses

2. Install the required Python packages:

   pip install -r requirements.txt

1. Fill in the configuration files with your credentials and settings:

   • producer_config.json
   • sentiment_analysis_config.json
   • telegram_bot_config.json

   Example producer_config.json:

   {
     "youtube_api_key": "YOUR_YOUTUBE_API_KEY",
     "playlist_id": "YOUR_PLAYLIST_ID",
     "kafka_bootstrap_servers": "YOUR_KAFKA_BOOTSTRAP_SERVERS",
     "kafka_topic": "youtube_comments"
   }

   Example sentiment_analysis_config.json:

   {
     "kafka_bootstrap_servers": "YOUR_KAFKA_BOOTSTRAP_SERVERS",
     "gemini_api_key": "YOUR_GEMINI_API_KEY",
     "input_topic": "youtube_comments",
     "output_topic": "youtube_sentiments"
   }

   Example telegram_bot_config.json:

   {
     "telegram_bot_token": "YOUR_TELEGRAM_BOT_TOKEN",
     "kafka_bootstrap_servers": "YOUR_KAFKA_BOOTSTRAP_SERVERS",
     "kafka_topic": "negative_youtube_sentiments",
     "telegram_channel_id": "YOUR_TELEGRAM_CHANNEL_ID"
   }

1. Start the Kafka server and create the necessary topics.

2. Run the following KSQL queries to create streams:

   CREATE STREAM YOUTUBE_SENTIMENTS_STREAM (
       ID VARCHAR,
       TEXT VARCHAR,
       STATUS VARCHAR,
       AUTHOR VARCHAR,
       PUBLISHEDAT VARCHAR,
       UPDATEDAT VARCHAR,
       VIDEOTITLE VARCHAR,
       SENTIMENT VARCHAR
   ) WITH (
       KAFKA_TOPIC='youtube_sentiments', 
       VALUE_FORMAT='JSON'
   );
   
   CREATE STREAM NEGATIVE_YOUTUBE_SENTIMENTS_STREAM AS 
   SELECT * 
   FROM YOUTUBE_SENTIMENTS_STREAM 
   WHERE SENTIMENT='negative';

3. Run each Python file separately:

   python producer.py
   python sentiment-analyses.py
   python telegram_bot.py

The project consists of the following files:

• LICENSE: Contains the project's license information.
• README.md: Provides an overview of the project, its purpose, and usage instructions.
• producer.py: Script for collecting YouTube comments and sending them to Kafka.
• producer_config.json: Configuration file for producer.py.
• sentiment-analyses.py: Script for performing sentiment analysis on YouTube comments using the Gemini API.
• sentiment_analysis_config.json: Configuration file for sentiment-analyses.py.
• telegram_bot.py: Script for sending negative comments to a Telegram channel.
• telegram_bot_config.json: Configuration file for telegram_bot.py.

This project is licensed under the MIT License. See the LICENSE file for details.