Skip to content

Latest commit

 

History

History
115 lines (76 loc) · 3.8 KB

README.md

File metadata and controls

115 lines (76 loc) · 3.8 KB
Raw

Kleio bot

https://twitter.com/KleioBot

A Twitter bot that:

  • Aggregates the last 50 tweets of each political party currently in the Greek parliament using the Twitter API

  • Uses Natural Language Processing with the help of a pre-trained model in the Greek language to parse sentences, group words with similar roots (lemmatisation) and dismiss common words (stop words)

  • Creates and posts a word cloud image on its Twitter account for each political party using the above data. This process is not always perfect and you might occasionally notice some duplicated words, or some verbs being conjugated incorrectly. This is due to the language processing model being used, and it will hopefully be improved over time

I have named it Kleio Bot after the muse Kleio from the Greek mythology who was ' the proclaimer, glorifier and celebrator of history, great deeds and accomplishments'.

It feels mundane to mention this, but for all intents and purposes this is a project just for fun and to explore the capabilities of technology when it interacts with a very rich and unique language such as Greek. This bot does not ' think' for itself, does not have political affiliations and does not take sides. It simply reproduces the data that it is given and presents it in a very specific and interesting way.

Prerequisites

Ubuntu (using apt)

curl -sSL https://raw.githubusercontent.com/python-poetry/poetry/master/get-poetry.py | python -
sudo apt install python3-pip python3-dev

MacOS (using brew)

brew install python@3.9 poetry

Windows

Use the above links to download the appropriate executables and install them.

Installation

Install dependencies

poetry install

Activate virtual environment shell

poetry shell

Create a Twitter app and generate the necessary credentials

In order for this bot to make use of the Twitter API, you must first create an app and generate the necessary consumer keys and access tokens. Refer to the Twitter API documentation on how to do that.

Environment variables

Create a .env file at the root of the project and add the consumer keys and access tokens generated for your app:

CONSUMER_KEY=...
CONSUMER_SECRET=...
ACCESS_TOKEN=...
ACCESS_TOKEN_SECRET=...

Running the bot manually

Warning! Running this application will actually post word clouds to the Twitter account you have linked using the environment variables above. To avoid this and just see the generated pictures, comment out the following line:

api.update_with_media(image_path, title)

TODO - I might make this adjustable using an environment variable

python main.py

Running the bot periodically

I am currently using cron to schedule the bot to run every day at 9:00 AM UTC (which translates to 11:00 AM Greece time). You can easily experiment using the same process.

You can edit your cron schedule with:

crontab -e

Assuming the project resides in a home folder of a user your_user with Python and poetry installed on the same box, then you can schedule it with adding to your crontab a line like the following:

0 9 * * * cd /home/your_user/kleio-bot && /home/your_user/.local/bin/poetry run python main.py

Refer to cron documentation about the necessary syntax to use.