MIT License • Code of Conduct • Contributing Guidelines • Security Policy

---

ScamShield: Fake Job Posting Detection Platform

ScamShield is an advanced, real-time AI/ML platform designed to protect job seekers from fraudulent employment opportunities. By actively auditing job descriptions, verifying URLs and corporate registrations, and analyzing recruiter contact channels, ScamShield calculates a comprehensive Trust Score (0-100%) before applicants submit their sensitive data.

---

🎥 Demo Video

Click the image above to watch a full demonstration of ScamShield in action.

---

✨ Key Features

• Real-Time Threat Detection: Instantaneous analysis of job postings to identify potential scams.
• Smart URL Scraping: Automatically fetches and populates job details (title, company, description) from major job boards for seamless analysis.
• 5-Pillar Trust Engine: Employs a multi-faceted verification pipeline incorporating machine learning and heuristic risk metrics.
• Premium Visual Dashboard: A sleek, responsive user interface featuring a dynamic horizontal pipeline tracker, theme toggling (Obsidian Dark / Alabaster Light), and detailed trust telemetry.

---

🛠️ System Architecture

ScamShield evaluates job postings using a 7-Stage Horizontal Verification Pipeline:

graph LR
A[Ingested] --> B[NLP Text]
B --> C[URL Scan]
C --> D[Company Check]
D --> E[Recruiter Check]
E --> F[Trust Score]
F --> G[Decision]

Loading

The system verifies credibility across 5 Trust Pillars:

1. 🧠 NLP Text Analysis:
   • Evaluates semantics using a high-performance machine learning text classifier.
   • Audits vocabulary density and detects clickbait signals (e.g., excessive exclamation marks).
   • Analyzes lexical complexity via the Flesch Reading Ease index to expose convoluted, machine-generated templates.
2. 🔗 URL Phishing Scan:
   • Inspects embedded links for secure SSL protocols (https:// vs http://).
   • Audits domains against suspicious free hosting providers (e.g., WordPress, Wix) and URL shorteners (e.g., bit.ly).
3. 🏢 Company Verification:
   • Validates description completeness to flag exceptionally brief or missing company profiles.
   • Audits company names against official corporate registry structures (identifying standard suffixes like Ltd, Inc, LLP, Pvt).
4. 🛡️ Recruiter Behavior Check:
   • Scans for embedded personal emails and phone numbers in the description text.
   • Distinguishes between public free email providers (e.g., @gmail.com) and validated corporate domains.
   • Cross-references job locations against historical regional fraud ratios.
   • Flags pressure tactics and urgency terms (e.g., "URGENT", "IMMEDIATE").
5. 🎯 Trust Score Engine:
   • Integrates machine learning probabilities and heuristic risk metrics to compute a credibility rating (0 to 100%).
   • 🟢 80% - 100%: High Trust (Verified Safe / Approved)
   • 🟡 50% - 79%: Neutral Risk (Verification Advised)
   • 🔴 Below 50%: Low Trust (Scam Vulnerability Alert / Application Blocked)

---

📈 Model Performance & Telemetry

Our core prediction models have been meticulously engineered to maximize prediction accuracy and minimize false-positive warning triggers.

• Text Classifier (NLP): Upgraded to TfidfVectorizer (10,000 max features) combined with an SGDClassifier(loss='log_loss'). Trained on balanced sets using SMOTE, filtering vocabulary noise and raising the individual text F1-score to 80.97%.
• Tabular Classifier: Utilizes a powerful RandomForestClassifier (100 estimators) to naturally model non-linear correlations (e.g., location threat ratios, character lengths, Flesch scores) without scaling bias, achieving an individual tabular F1-score of 56.32%.
• Ensemble Fusion (Soft Voting): Combines model outputs using a weighted soft-voting algorithm (0.5 * Text Risk + 0.5 * Tabular Risk) combined with a heuristic check penalty.

Overall Ensemble Performance:

• Accuracy: 98.11%
• Precision: 89.86% (Significantly reducing false alarm flags)
• Recall: 80.52%
• F1 Score: 84.93%

---

🚀 Quick Start Guide

1. Prerequisites

Ensure you have Python 3.12+ installed. Install the required dependencies:

pip install -r requirements.txt

(Or install manually: pip install flask pandas numpy scikit-learn imbalanced-learn joblib textstat beautifulsoup4 nltk python-docx)

2. Train the Models

Train and serialize the TF-IDF vectorizers and Random Forest classifiers on the provided balanced dataset.

python train_model.py

This step outputs accuracy metrics and saves the binary model files (tfidf_vectorizer.pkl, clf_log.pkl, clf_num.pkl, numeric_features.pkl) into the models/ directory.

3. Launch the Application

Start the Flask development server:

python app.py

Navigate to http://127.0.0.1:5000 in your web browser to access the ScamShield dashboard.

🤝 Contributing

We welcome contributions to ScamShield! Please see our Contributing Guidelines for more details on how to get started, report bugs, or submit pull requests.

📄 License

This project is licensed under the MIT License.