Project Overview

Project Summary:

'FishingPhishing' is an application that provides smishing prevention, education, and a community for the victims. It determines the risk level of texts by analyzing suspicious URLs using Machine Learning model. The application also provides "Fake Smishing", where fake messages, created by Generative AI are randomly sent to the users. It is to help the users increase their ability to discern the smishing messages. If users click on a URL in these fake messages, FishingPhishing leads users to 'Smishing Education' tab. Also, users can share their experiences and fight against the smishing crime together through community function.

Identifying the Challenge

The Social Problem:

97% of Korean adults use smartphones to handle important personal information. As reliance on smart devices increases, it is expected that even more sensitive personal data will be managed through these devices in the future. This implies that smishing crimes will inflict even more devastating harm on individuals. Furthermore, smishing serves as a starting point for other forms of information breaches.

Therefore, by preventing smishing, we can protect personal information and prevent the proliferation of information breaches stemming from smishing.

Innovation and Uniqueness

Why Our Project Stands Out:

The differentiating factors of FishingPhishing are "joint response" and "education." While existing apps have focused on detecting smishing messages, FishingPhishing concentrates on preventing smishing and aiding in post-incident response through these two features.

Through community features, users can share specific information about smishing methods and coping strategies, enabling users to handle smishing wisely.

The fake phishing feature serves as a shock therapy to raise users' awareness of smishing, and quizzes help users learn about smishing without getting bored.

Insights and Development

Learning Journey:

Our team focused on ensuring that all team members understood and participated in the project sufficiently. It wasn't easy to bring together ideas from individuals with different backgrounds, but by leveraging each person's abilities, we were able to develop ideas together and understand each other's efforts. Furthermore, the diversity and comprehensive perspective ultimately enriched the project.

Development Process:

We started a project to combat smishing crimes. Through discussions, we sought ways to solve smishing problems and differentiate our project from other services. Eventually, we concluded to emphasize joint response and education features through 'reporting' and 'community' tab.

Using Python's 'Jupyter Notebook', we developed a machine learning model after analyzing 'Design and Implementation of a Malicious URL Prediction System Using Multiple Machine Learning Algorithms'. Despite challenges in data collection and utilization, we resolved issues together by dividing roles and discussing preprocessing methods.