18649594. PRIVACY FILTERS AND ODOMETERS FOR DEEP LEARNING simplified abstract (Microsoft Technology Licensing, LLC)

PRIVACY FILTERS AND ODOMETERS FOR DEEP LEARNING

Organization Name

Microsoft Technology Licensing, LLC

Inventor(s)

[[:Category:Mathias François Roger L�cuyer of New York NY (US)|Mathias François Roger L�cuyer of New York NY (US)]][[Category:Mathias François Roger L�cuyer of New York NY (US)]]

PRIVACY FILTERS AND ODOMETERS FOR DEEP LEARNING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18649594 titled 'PRIVACY FILTERS AND ODOMETERS FOR DEEP LEARNING

The patent application discusses devices, systems, and methods for improving phishing webpage content detection through the use of a nested privacy filter architecture and a privacy odometer.

• Instantiating an odometer with a nested privacy filter architecture
• Training a DL model with privacy filters of different sizes
• Maintaining a running total of privacy loss budget consumed during training
• Returning the size of the smallest privacy filter bigger than the total privacy loss budget consumed

Potential Applications: - Enhancing cybersecurity measures for detecting phishing websites - Improving privacy protection for users browsing the internet

Problems Solved: - Enhancing the detection of malicious phishing webpage content - Providing better privacy protection for internet users

Benefits: - Increased accuracy in detecting phishing websites - Enhanced privacy protection for users online

Commercial Applications: Title: Advanced Phishing Detection System This technology can be used by cybersecurity companies to enhance their phishing detection capabilities, providing a more secure browsing experience for users. Market implications include increased demand for advanced cybersecurity solutions.

Questions about Phishing Detection: 1. How does the nested privacy filter architecture improve phishing webpage content detection?

  - The nested privacy filter architecture allows for the training of a DL model with privacy filters of different sizes, enhancing the accuracy of phishing webpage content detection.

2. What are the potential benefits of using a privacy odometer in detecting phishing websites?

  - The privacy odometer helps maintain a running total of privacy loss budget consumed during training, providing a measure of privacy protection for users.

Original Abstract Submitted

Generally discussed herein are devices, systems, and methods for improving phishing webpage content detection. A method can include instantiating an odometer with a nested privacy filter architecture, the nested privacy filter including privacy filters of different, increasing sizes, training a DL model, maintaining, during training and by a privacy odometer that operates using the nested privacy filter, a running total of privacy loss budget consumed by the training, and responsive to a query for the total privacy loss budget consumed, returning, by the odometer, a size of a smallest privacy filter of the nested privacy filters that is bigger than the running total of the privacy loss budget.