AWS intends to enhance its collaborative data-sharing service, AWS Clean Rooms, by adding machine learning capabilities. This new feature will enable enterprises and their partners to apply machine learning to shared datasets without having to exchange raw data, thereby generating predictive insights.
By creating lookalike segments of data resembling the original raw dataset, AWS Clean Rooms ML lookalike modeling will allow users to train custom models using their data and invite partners to collaborate and generate expanded sets of similar records while safeguarding underlying data.
The machine learning functionalities of AWS Clean Rooms, available as a standalone offering and as part of AWS for Advertising and Marketing, will be accessible in various regions.
AWS Clean Rooms, released at re:Invent last year, can be accessed via the AWS Management Console, providing options to choose data sets for sharing and configure permissions for participants. The data sets shared within the clean room are encrypted, eliminating the need to move out of the AWS environment or be loaded into another platform.
Both the enterprises and their partners can run queries on the datasets, with a broad set of configurable data access controls allowing for customization of restrictions on the queries run by each clean room participant.
Copyright © 2023 IDG Communications, Inc.
