Overview / Usage

The bot would scold the user who use offensive language and give warnings to the user. If the user still doesn't mend the ways, the bot would kick the user out of the server. The bot basically checks if any person while using offensive words, crossed a threshold. If the user crossed a threshold, then they are kicked from the user and the kicked user also gets a mail that they were kicked due to inappropriate behavior.

Methodology / Approach

I first collected a long list of cuss words, discriminatory and racial slangs. First we created a bot from discord, copied its credentials of oAuth2 and permissions accordingly in our local server and made the files and code accordingly. After making the bot online, it becomes active and can function now. Now if any user send any message on discord, then I check if any of the listed word is present in the message and I mark the user in our local storage indicating the offensive behavior from that particular user to further check any kind of offensive behavior in future. If the message contains any listed word, then the bot send warnings to the user to not to use such words. If the user continues to use the offensive language then the bot kicks the user off the server. We implemented data structure trie to store the offensive words to speed up message analyses and fast decision by the bot. We implemented all of this in Python using discord.py library to ensure smooth functioning of the bot. We used notivize's mailing service to send mails to the kicked user. I then pushed the hosted our bot on heroku to be used globally 24/7 anywhere.

Technologies Used

Python, Machine Learning, scikit learn intelex, String Matching