Bot parser of a news site, with post generation through GPT
This is an automated bot that parses news from the Ukrainian Truth website (pravda.com.ua), filters them by keywords and publication time (not older than 2 hours), generates brief summaries using ChatGPT (GPT-4), and sends them to administrators for moderation. After approval, the bot publishes the news in a Telegram channel.
Main functions:
Automatic news parsing every hour
Filtering by keywords
Generating brief summaries using GPT-4
Moderation through buttons (Approve / Reject)
Auto-publishing in the Telegram channel after approval
Protection against duplicates (does not process already sent links)
Technologies used:
Python 3.11+ (asynchronous code, asyncio, aiohttp)
Aiogram 3.x (Telegram Bot API)
BeautifulSoup4 (HTML parsing)
APScheduler (task scheduler)
GPT-4 API (via g4f for content generation)
Logging (logging the bot's work)
Why is this project useful?
Automates routine work – no need to manually search for news.
Saves time – GPT-4 writes brief summaries by itself.
Flexibility – keywords and sources can be changed.
Scalability – other news websites can be added.
Stack: Python, Aiogram, GPT-4, BeautifulSoup, APScheduler
Main functions:
Automatic news parsing every hour
Filtering by keywords
Generating brief summaries using GPT-4
Moderation through buttons (Approve / Reject)
Auto-publishing in the Telegram channel after approval
Protection against duplicates (does not process already sent links)
Technologies used:
Python 3.11+ (asynchronous code, asyncio, aiohttp)
Aiogram 3.x (Telegram Bot API)
BeautifulSoup4 (HTML parsing)
APScheduler (task scheduler)
GPT-4 API (via g4f for content generation)
Logging (logging the bot's work)
Why is this project useful?
Automates routine work – no need to manually search for news.
Saves time – GPT-4 writes brief summaries by itself.
Flexibility – keywords and sources can be changed.
Scalability – other news websites can be added.
Stack: Python, Aiogram, GPT-4, BeautifulSoup, APScheduler