Articles on: Getting Started

ASN-based blocking & machine learning model

Overview



Tapper has implemented an advanced ASN-based blocking system that combines real-time data analysis and machine learning to differentiate between good and bad Autonomous System Numbers (ASNs). This ensures that ad fraud and bot-driven traffic are blocked while maintaining access for legitimate users.

This guide explains how the system works, the factors influencing ASN classification, and how you can fine-tune the settings based on your needs.

How Tapper Blocks Data Center Traffic More Effectively



By integrating IPData’s ASN intelligence and Tapper’s machine learning models, our system can:

✅ Automatically detect and block fraudulent ASNs in real-time.
✅ Prevent unnecessary blocking of legitimate cloud-based users who may be using hosting services.
✅ Allow manual rule creation to override AI-driven classifications.
✅ Continuously improve blocking accuracy with user feedback and iterative updates.

1. How Tapper Categorizes ASNs



Tapper classifies ASNs into Good and Bad based on a range of data points.

Factors Considered for ASN Classification



Head 1Head 2
Usage TypeHosting providers and data centers are more likely to generate bot traffic.
Upstream/Downstream ASNSASNs frequently connected to known malicious networks are considered high risk.
IP Range DensityASNs with a large number of IPs are often used for bot traffic.
Country of RegistrationCertain regions have a higher concentration of fraudulent activity.
ASN AgeNewly assigned ASNs are more likely to be used for fraud.
Traffic BehaviorSudden spikes in low-quality clicks indicate bot-driven traffic.

## 2. Machine Learning Model for ASN Classification

Tapper’s machine learning engine enhances ASN blocking by analyzing behavioral data from multiple sources.

How the Model Works:



Data Collection – IPData API is used to fetch real-time ASN details.
Feature Extraction – Usage type, peer relationships, and fraud history are used to classify ASNs.
Binary Classification Model – Each ASN is assigned a confidence score (0-1) indicating its likelihood of fraud.
Blocking Thresholds – Based on the confidence score, ASNs are either:

✅ Allowed (Low-risk ASNs)
⚠️ Flagged for Review (Medium-risk ASNs)
❌ Blocked Automatically (High-risk ASNs)

Customizing Your Blocking Rules



Tapper provides a Real-Time Configuration Panel where you can:

Set your own blocking threshold (e.g., block ASNs with a fraud score above 0.8).
Whitelist trusted ASNs (e.g., cloud service providers).

Manually block specific ASNs based on campaign needs.

3. Implementing ASN Blocking in Your Tapper Account



Step 1: Enable ASN Blocking in Your Dashboard



Log in to your Tapper account.
Navigate to Settings > Traffic Protection.
Toggle on “Block Suspicious ASNs.”
Set your Confidence Score Threshold (default: 0.8).

Step 2: Review and Manage Blocked ASNs



Go to Analytics > ASN Reports to view flagged ASNs.
Click on an ASN for detailed risk analysis.
Use the Allowlist or Blocklist feature to manually adjust settings.

Step 3: Monitor Performance & Optimize



Track blocked vs. allowed traffic in your Fraud Prevention Dashboard.
Adjust blocking sensitivity based on real-time performance.
Submit feedback on incorrect classifications to improve machine learning accuracy.

4. Continuous Model Improvements



Tapper’s ASN blocking system continuously improves through:

🔄 Real-world feedback from users to refine accuracy.
📊 Data updates from IPData & security feeds to keep fraud lists current.
🤖 Machine learning retraining every 30 days for enhanced precision.

5. Common Questions



Q: Will this block legitimate users who use cloud-based hosting?


No, Tapper differentiates between consumer traffic and data center traffic to avoid blocking valid users. You can always whitelist trusted ASNs.

Q: Can I create my own ASN blocking rules?


Yes! You can manually override any automated decision using the ASN Allowlist or Blocklist in the dashboard.

Q: How often is the blocking list updated?


Our system updates every 24 hours based on new fraud trends and user feedback.

Conclusion



Tapper’s ASN-based blocking system ensures that your ads only reach real users, not bots. By using a combination of real-time IP intelligence, machine learning, and customizable rules, Tapper maximizes protection while minimizing false positives.

For any custom requirements or technical assistance, feel free to reach out to our support team at support@tapper.ai.

Updated on: 03/03/2025

Was this article helpful?

Share your feedback

Cancel

Thank you!