Inspiration
I remember when I had to check into a workplace for the first time. Among many things, they were very strict about data security, especially in regards to phishing emails. For many companies, phishing can be one of the greatest threats to their bottom line. Therefore, I was inspired to create an AI solution to this problem.
What it does
Our program takes in an email, converts it to a data file, and runs it through our AI classification algorithm. We detect if the email has any urgent or suspicious language, and assign it as a high risk email or a low risk email. Then, it is run through classification to finally determine if the email is spam or safe.
How we built it
We started by collecting and processing data, which contained spam and safe emails. Then, we created a sentiment detector to figure out if the language seemed urgent. Using that evaluation and the evaluation of other non-machine learning algorithms like checking for spelling errors, we assign a label of high or low risk to an email. Then, we initialised a supervised learning model, training it on our labeled data. Finally, we created a website to allow users to input emails and returns if the email is spam or not
Log in or sign up for Devpost to join the conversation.