PhishHook

Inspiration

Phishing is the most common cyber attack to business and people. Phishing emails will prey on human behavior, fostering a sense of urgency or reward to get the user to do something that compromises their information. We are all just fish in the sea, and it's time to stop getting hooked.

What it does

Our Website, PhishHook, checks the email contents itself. Our algorithm using natural language processing, checking for common characteristics: sense of urgency, promise of a reward, grammar, etc. to detect a phishing attempt.

How we built it

We built the front-end using streamlit. For our phishing detection algorithm, we trained a machine-learnning model on several data sets including phishing emails dataset from Berkeley, wordcloud (described unique words often found in phishing emails and the rate at which they appear). We also check of tone, grammar, and keywords.

Challenges we ran into

Everything front-end. Initially, the app was intended to be a chrome extension, but we could not get Gmail API and a python script without running into Google security policies. Next, running a python script on a website was more difficult than we thought, and we learned streamlist basics to do so.