Inspiration

What it does

Out product Phishing Defender basically checks all the links present in on a web page or website and display whether a link is phishing or legitimate. This is done by passing the links to a trained machine learning model which has potential accuracy and precision of more than 95%.

How we built it

  • Phishing Defender was built using Python with Beautiful Soup for web scraping and Flask for the web application. The core protection comes from a machine learning model trained to detect phishing links. Our teamwork and focus on cyber security resulted in a straightforward yet effective solution for a safer online experience.

Challenges we ran into

  • The first and major challenge was that our group was freshers when we thought of the idea, so we started from scratch just with the idea, it was a long journey for us to complete this project.
  • The second major challenge was to integrate the different components of project with each other, like getting POST request from machine learning model to then sending a GET request to flask server.
  • The third challenge we faced as a team was to increase the accuracy of model as it took many alterations

Accomplishments that we're proud of

  • Effective Phishing Detection: We take pride in developing a project that does not require the user to visit the phishing link at all.

  • Seamless Integration : Achieving seamless integration of web scraping with Beautiful Soup and Flask has allowed us to provide users with a user-friendly and efficient tool.

  • Positive User Impact: With data bombarding every day, and newer attempts to fraud users, it has become easy for unethical practitioners to fraud us easily, moreover we do not have time to check each link whether it is trustworthy or not, so our project is a life-saver!

What we learned

  • Collaboration and Communication: The most important thing we learned as a team was the importance of effective communication and collaboration within a team, especially when dealing with interdisciplinary tasks involving machine learning, web development, and cybersecurity.

  • Machine Learning Tuning: Fine-tuning a machine learning model for phishing detection taught us the intricacies of parameter adjustments and feature selection, enhancing our skills in model optimization.

  • Web Scraping Challenges: Overcoming challenges in web scraping broadened our understanding of handling diverse HTML structures, ensuring the reliability of data extraction across various websites.

  • Real-World Application: Making Aware to our surroundings and ourselves about cybersecurity importance and consequences of data loss

What's next for Phishing Defender

  • We have a plan to make it more user-friendly more seamless, and spreading the word about online security
  • Continue to update and upgrade the machine learning model as newer and newer phishing links will emerge but we must fight all those
  • Making our plugin more interactive by introducing AI into it
Share this project:

Updates