Inspiration

Doxing, or a release of identifying information, has become an increasing threat in this digital age. According to Discord, doxxing, harassment, and threatening behaviour account for over 30% of their received reports.

What it does

Doxcord parses through messages and scans them for identifying information. If a doxing attempt has been detected, then the message is deleted and a copy of the message is sent to the server's admin team for further investigation.

How we built it

Doxcord is built using a pre-trained NLP library, SpaCy to help with the detection of personal information. The Discord API was used to develop the Discord Client, and handle reading, writing, and deleting messages. The bot is hosted using Google Cloud.

Challenges we ran into

We ran into a number of challenges while implementing the program, such as when multiple messages were being sent at once or trying to detect malicious intent of a message.

Accomplishments that we're proud of

Overall, we are proud that we were able to develop a MVP in such a short amount of time, using technologies that we were not the most familiar with, and a language that one of us does not use regularly!

What we learned

As this was one of our first hackathons we learned a lot about building a project in a short period of time. We also learned how to sift through technical documentation, and Google effectively.

What's next for Doxscord

The group wishes to make a number of improvements to the product. These include: adding an admin flagging event for false positives to strengthen the model, handling edge cases to avoid bypassing the bot's features, and developing a website so that users can learn more about the bot and add it to their own servers.

Share this project:

Updates