Skip to content

Summary of the 3 most interesting AI Research Papers daily on your own telegram channel/group

Notifications You must be signed in to change notification settings

robot-bulls/Daily-AI-Research-Paper-Summary-Telegram

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI Research Paper Scraper Telegram

The AI Research Paper Scraper Telegram is a Python script designed to fetch recent submissions from arXiv's Computer Science category, extract and summarize text from the linked PDFs, and then send these summaries to the specified Telegram channel. You can view the results in this Telegram channel: @robotbulls_ai_papers.

This script get the 3 best papers daily and posts it on this Telegram channel. You can change this number in the script.

Right now it gets the 3 most interesting papers for computer science but you can change the subject to math, physics etc.

Installation

pip install -r requirements.txt

Configuration

You will need to configure the following variables within config.json:

  • api_key: Your OpenAI API key.
  • token: Your Telegram bot token. You can get this by messaging the BotFather on Telegram.
  • channel_id: Your Telegram channel ID. Make sure that the bot is an admin of your Telegram channel.

Usage

Run the script from your terminal:

python main.py

This will execute the script, which performs the following steps:

  1. Fetch Recent Papers: Fetches papers submitted to the arXiv Computer Science category the previous day.
  2. Reduce Selection: Iteratively refines the selection of papers until a manageable number is obtained for summarization.
  3. Extract and Summarize: Extracts text from each paper's PDF and generates a summary.
  4. Send to Telegram: sends each summary to the specified Telegram channel.

About

Summary of the 3 most interesting AI Research Papers daily on your own telegram channel/group

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages