FishThePhish

Inspiration

with the current state of affairs, there has been an increase in the vulnerability of desperate people looking for medical advice concerning COVID, our application combats this by considering multiple red-flags such as grammar, email-verification, and context. In order to filter out adversarial emails looking to exploit people's crave for information regarding their health, we are developing a phishing detection app to alert users to potential misinformation

What it does

It tells people whether or not their email messages seem strange.

How I built it

I scrapped all emails I have read in my personal email in order to identify what constitutes a normal email given the polarity and grammatical structure scores. I feed my training data into an autoencoder in order to learn a good generality of a normal email. I displayed my results on a very basic HTML page.

Challenges I ran into

This was my first time testing out the autoencoder model and doing any type of frontend development.

Accomplishments that I'm proud of

I was able to develop my own model from scratch and interface it on a simple web app for the first time.

What I learned

I learned about HTML

What's next for FishThePhish

Update the frontend with a react interface to make it look better. I also want to make some modifications to the model in order to account for other factors in an email.

Built With

• flask
• html
• juypter