ArXiv Today

ArXiv Today: Get arXiv daily papers right in your Lark (飞书) via bot.

ArXivToday-Lark is a lightweight tool that automates the process of fetching the latest papers from arXiv and delivers them directly to your Lark group chats using a custom bot. Designed for research enthusiasts and academic professionals, this project simplifies daily paper discovery with customizable features, seamless integration, and extendable functionality.

Key highlights include automated scheduling, support for LLM-based paper filtering, summary translation, and influence prediction (in development). Whether you’re exploring cutting-edge research or curating papers for your team, ArXivToday-Lark makes it easy and efficient to stay updated.

Demo

To Do

• Use LLMs for more accurate paper filtering.

• Use LLMs to translate paper abstracts.

• Predict paper impact using LLMs.

  Zhao P, Xing Q, Dou K, et al. From Words to Worth: Newborn Article Impact Prediction with LLM[J]. arXiv preprint arXiv:2408.03934, 2024.

Usage

Prerequisite

1. Clone this repository.

   git clone https://github.com/InfinityUniverse0/ArXivToday-Lark.git

2. Install the required Python packages.

   cd ArXivToday-Lark
   pip install -r requirements.txt

Deployment

In Lark, add a Custom Bot to a group chat. Deploy and run this project to fetch the latest relevant papers from arXiv daily and push them to the group via the bot.

Add a Lark Custom Bot

Follow the steps in this guide to add a custom bot to your group chat in Lark.

Set Up Lark Message Card Templates

Refer to this guide for detailed steps on setting up message card templates in Lark.

The message card template used in the Demo can be directly imported from ArXivToday.card and applied in Lark.

Configure Script Parameters

In config.py, modify the following parameters based on the results of the previous steps:

1. Webhook URL of the Lark bot.
2. ID and version number of the Lark message card template.

Adjust these settings according to your specific setup.

Run the Script

Run the script using Python:

python main.py

To run the script periodically, you can use the crontab command in Linux or the schedule library.

Run Periodically with crontab

Requires a Linux system

For example, to fetch arXiv papers and push them via the Lark bot at 12:24 PM every weekday, follow these steps:

1. Open the crontab editor with the following command:

   crontab -e

2. Add the following line and save it:

   24 14 * * 1-5 /absolute/path/to/your/python/interpreter /absolute/path/to/ArXivToday-Lark/main.py

Note

⚠️ Ensure to provide absolute paths for both the Python interpreter and the script.

3. Verify the crontab task setup with this command:

   crontab -l

Run Periodically with the schedule Library

1. Install the dependency:

   pip install schedule

2. Uncomment the following section in main.py and modify it as needed:

   ### Uncomment the following code to use `schedule` to run the task periodically ###
   import time
   import schedule
   # Schedule the task to run every day at 10:17
   schedule.every().day.at("10:17").do(task)  # TODO: Change the time for your own need
   while True:
       schedule.run_pending()
       time.sleep(1)

Extension

This project can be extended to meet custom requirements. For instance:

• You can design your own message card styles or use other message types.
• You can integrate a Lark App Bot (might require additional permissions) to implement more complex workflows.

License

This project is under the GPL-3.0 License.

Contact

For any questions, suggestions, or feedback, feel free to reach out:

• Email: [email protected]
• GitHub Issues: Issues Page

Feel free to contribute, report issues, or suggest improvements!