Adobe Inc.
Machine learning techniques for adaptation of initial workspace

Abstract:

An initial workspace recommendation engine recommends a user interface component available for inclusion in an initial workspace of a newly installed application instance. In some cases, a classifier included in the engine classifies the available user interface component based on data describing multiple previously installed instances of the application. For example, the multiple instances are installed on multiple user devices, and are associated with user features describing the installed instances. In addition, the initial workspace recommendation engine receives a set of new user features associated with the newly installed application instance. The classifier classifies the available user interface component based on a similarity between the new user features and the user features describing the installed instances. Based on the classification, the initial workspace recommendation engine recommends the available user interface component for inclusion in or omission from the initial workspace of the newly installed instance.