The initial Idea for the SpamPo Bot came from our desire to apply our knowledge with Machine Learning and Artificial Intelligence to produce a solution to Discord spambots. As avid Discord users, spambots were often common occurrences in larger and more prominent servers of which we were members; thus, by conceptualizing a solution through the consideration of Ignite Hacks main topics – Machine Learning and Artificial Intelligence – we determined that teaching a bot through Machine Learning models to identify spambots would aid in the process of detecting and handling such bots for future Discord users.
What it does
The SpamPo Bot, first and foremost, logs any confirmed scam/spambots; it also records the message history between the bot and a user and by doing so, the bot uses a neural network (LSTM architecture) to predict and identify scam/spambots or users. The new data gained from recording the message history is used to train the Neural Network of the bot. The bot also provides a list of potentially problematic users upon request when the user indicates as so.
How we built it
In order to build the SpamPo Bot, we had to incorporate Machine Learning Methods learned from the Ignition Hacks Workshops as well as the Discord Python API and Library; this means that we had to use Discord to test the efficacy of the bot and its functionality. We also had to use our knowledge of Python to construct the appropriate code to provide the bot with the proper functionalities as intended by the purpose of the bot.
Challenges we ran into
As a team, it was difficult to integrate the use of Github as the primary source-code host where we would construct and edit code; resources similar to Github are a novel concept as much of the team experiences were as individual programmers and developers, often subverting the need for a collaborative medium. Another difficulty we face was grasping the concept of Machine Learning and actualize the theory of Machine Learning through the SpamPo Bot; specifically, one of the difficulties was the lack of knowledge regarding the syntax related to Machine Learning Methods and the actual implementation of the syntax.
Accomplishments that we're proud of
As a team, we are proud that we have attained an understanding of how to use GitHub as a platform to construct and edit code. We are also proud of the fact that we have created a functional Discord bot that efficiently and accurately executes its intended purpose of detecting and handling spambots. Furthermore, we're proud of our new understanding of Machine Learning and its implementation in an actual application such as the SpamPo Bot.
What we learned
Over the duration of Ignition Hacks, our team has acquired a plethora of new knowledge relating to Machine Learning and Artificial Intelligence. We have also learned more about possible careers in Computer Science and Technology. Specifically, an aspect of the content learned for this Hackathon is in regards to utilizing the Discord Python API and understanding the syntax and functions related to the Discord Python Library.
What's next for SpamPo Bot Development for the Discord App
Through the development of the SpamPo Bot, specifically for Ignite Hacks, we hope that we can further expand the functionalities of the bot, in the future, to aid more Discord users in detecting and handling undesired spambots. We also hope to help many discord servers provide a pleasant experience to their users, in the absence of unsolicited advertisements and spam.
Log in or sign up for Devpost to join the conversation.