-
UI for PhishingNet homepage
-
UI for PhishingNet analysis tool
Inspiration
I got the idea for this project while I was working on getting certified in cybersecurity. I thought it would be a good idea to build something that implements some of the concepts that I have learned during the process.
What it does
PhishingNet's goal is to check if the text or email that you have, which you believe to be potentially suspicious, is a phishing attempt. It uses multiple datasets and a complex framework to make sure you are safe and not exposed to any malicious data.
How I built it
I built this project originally with just one data set. However, as the process went on, I realized that this was not enough, as the model was not at the level that I wanted it to be. Due to this, I added another, more recent dataset, along with some more complex feature engineering to create the best model I can.
Challenges I ran into
The biggest and most frustrating challenge I ran into was the model not living up to the expectation. There was a lot of feature engineering I had to do to create the most accurate model. Even after merging multiple datasets to train on, it was still quite faulty, not recognizing some key words that would indicate a scam. After many hours of manual correction, the final product is something that I am very proud of.
What I've learned
This was my first project dealing with natural language processing. This was something that I have always found interesting, and I know that this is something that every scientist should be comfortable with. I have learned more about development servers and how, for a project like this, it is not necessary to have a large frontend folder. Overall, I have learned more about web development as a whole. I am thankful to get another project under my belt and, more importantly, I am excited to build more!
Log in or sign up for Devpost to join the conversation.