Cracking the Code: Detecting Fake Job Posts with AI

Cracking the Code: Detecting
Fake Job Postings Using AI
A Machine Learning-Based Approach
to Secure Online Recruitment

Introduction
• • What are fake job postings?
• • Why are they dangerous?

Problem Statement
• • Increase in job scams online
• • Lack of robust detection systems
• • Manual reporting is slow

Objective
• • Build a system to classify job postings as fake
or genuine using ML

Literature Review
• • Summary of 3-4 research papers
• • Gaps identified in existing studies

Dataset Used
• • Source: Kaggle
• • ~17,000 entries
• • Fields: title, description, location, salary, etc.

Data Preprocessing
• • Null value handling
• • Tokenization & Lemmatization
• • TF-IDF vectorization

Model Selection
• • Algorithms: Logistic Regression, Random
Forest, BERT
• • Comparison of accuracy

Evaluation Metrics
• • Accuracy, Precision, Recall, F1-score
• • Confusion Matrix

Results
• • Best model: BERT (93.5% Accuracy)
• • ROC Curve

System Architecture
• • Data Ingestion → Preprocessing →
Prediction → User Interface

Demo Snapshots
• • UI screenshots or system mockups

Limitations
• • Generalizability
• • Lack of real-time API

Future Work
• • Email integration
• • Real-time job portal scraping

Conclusion
• • Need for secure recruitment
• • ML can automate scam detection

Q&A
• • Questions and Answers

More Related Content