18158299. PRIVACY ENHANCED MACHINE LEARNING OVER GRAPH DATA simplified abstract (International Business Machines Corporation)

PRIVACY ENHANCED MACHINE LEARNING OVER GRAPH DATA

Organization Name

International Business Machines Corporation

Inventor(s)

Ambrish Rawat of Dublin (IE)

Naoise Holohan of Maynooth (IE)

Heiko H. Ludwig of San Francisco CA (US)

Ehsan Degan of Rancho Santa Margarita CA (US)

Nathalie Baracaldo Angel of San Jose CA (US)

Alan Jonathan King of South Salem NY (US)

Swanand Ravindra Kadhe of San Jose CA (US)

Yi Zhou of San Jose CA (US)

Keith Coleman Houck of Rye NY (US)

Mark Purcell of Naas (IE)

Giulio Zizzo of Dublin (IE)

Nir Drucker of Zichron Yaakov (IL)

Hayim Shaul of Kfar Saba (IL)

Eyal Kushnir of Kfar Vradim (IL)

Lam Minh Nguyen of Ossining NY (US)

PRIVACY ENHANCED MACHINE LEARNING OVER GRAPH DATA - A simplified explanation of the abstract

This abstract first appeared for US patent application 18158299 titled 'PRIVACY ENHANCED MACHINE LEARNING OVER GRAPH DATA

The abstract of this patent application describes a process for privacy-enhanced machine learning and inference, where a system modifies access to private data in a graph database to generate predictions in response to queries without directly exposing the private information.

• The system includes a processing component that generates an access rule to modify access to private data in a graph database.
• A sampling component executes a random walk for sampling data in the graph database while following the access rule.
• An inference component generates predictions based on the sampled data in response to queries, avoiding direct exposure of private information.

Potential Applications: - This technology can be applied in healthcare for analyzing sensitive patient data while maintaining privacy. - It can be used in financial services for making predictions based on confidential customer information. - Privacy-enhanced machine learning can also be valuable in legal and law enforcement applications for analyzing sensitive case data.

Problems Solved: - Protects sensitive information while allowing for machine learning and inference processes. - Ensures privacy compliance in data analysis and prediction tasks. - Reduces the risk of exposing private data during machine learning operations.

Benefits: - Enables organizations to leverage sensitive data for predictive analytics without compromising privacy. - Enhances data security and confidentiality in machine learning applications. - Facilitates the development of privacy-preserving algorithms for various industries.

Commercial Applications: Privacy-enhanced machine learning technology can be utilized in industries such as healthcare, finance, legal, and law enforcement for secure data analysis and prediction tasks, ensuring compliance with privacy regulations and safeguarding sensitive information.

Questions about Privacy-Enhanced Machine Learning: 1. How does this technology ensure the privacy of sensitive data during the machine learning process? 2. What are the key advantages of using privacy-enhanced machine learning in various industries?

Frequently Updated Research: Stay informed about the latest advancements in privacy-enhanced machine learning and inference by following research publications and conferences in the fields of machine learning, data privacy, and artificial intelligence.

Original Abstract Submitted

One or more systems, devices, computer program products and/or computer-implemented methods of use provided herein relate to a process for privacy-enhanced machine learning and inference. A system can comprise a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory, wherein the computer executable components can comprise a processing component that generates an access rule that modifies access to first data of a graph database, wherein the first data comprises first party information identified as private, a sampling component that executes a random walk for sampling a first graph of the graph database while employing the access rule, wherein the first graph comprises the first data, and an inference component that, based on the sampling, generates a prediction in response to a query, wherein the inference component avoids directly exposing the first party information in the prediction.