Inspiration With the rise in online job scams, I wanted to build a tool that helps job seekers identify fraudulent job postings quickly and easily. Many people fall victim to fake listings, so I aimed to create a practical solution using machine learning.

What I Learned Through this project, I learned how to handle real-world datasets, perform text preprocessing, and develop a classification model using XGBoost. I also gained experience in building an interactive dashboard to make predictions accessible and understandable for all users.

How I built it Data: Collected and cleaned a dataset of job postings, labeled as real or fraudulent.

Model: Trained an XGBoost classifier to predict scam probability.

App: Developed a Streamlit app where users can upload job listings as CSV files and instantly detect scams.

Dashboard: Added visualizations like fraud probability histograms, pie charts, top suspicious jobs, and location-based filters.

Challenges The main challenge was dealing with imbalanced data, as genuine postings greatly outnumber scams. I also focused on making the dashboard intuitive and easy to use for everyone.

Built With

Share this project:

Updates