A full-stack machine learning web application that detects fake job postings in real time using Natural Language Processing (NLP) and supervised learning techniques.

This project was built as part of the Infosys Springboard program to demonstrate end-to-end ML workflows — from data preprocessing and model training to web deployment and analytics.

• ~98% accuracy in fake job classification using TF-IDF + Logistic Regression
• Real-time prediction of job authenticity via a web interface
• Interactive analytics dashboard to visualize prediction trends
• Secure admin panel for managing data and monitoring model behavior
• Flag suspicious job posts for further review
• Export predictions and logs in CSV format
• One-click model retraining with updated datasets

Backend & ML

• Python
• Flask
• scikit-learn
• TF-IDF Vectorizer + Logistic Regression

Database

• SQLite

Frontend & Visualization

• HTML, CSS
• Chart.js

pip install -r requirements.txt
python app.py

Then open:

• User Interface: http://127.0.0.1:5000
• Admin Panel: http://127.0.0.1:5000/admin_login

Admin Credentials (Demo):

Username: admin
Password: admin123

• Trained on real-world job posting datasets
• Optimized preprocessing for imbalanced text data
• Designed with scalability and retraining in mind
• Demonstrates practical use of ML in recruitment fraud detection

• Job portals & recruitment platforms
• HR teams screening suspicious postings
• Educational demonstration of applied NLP & ML