Machine learning for adaptive user interfaces

Abstract

In this paper we examine the growing interest in personalized user interfaces and explore the potential of machine learning in meeting that need. We briefly progress in developing fielded applications of machine learning, then consider some characteristics of adaptive user interfaces that distinguish them from more traditional applications. After this, we consider some examples of adaptive interfaces that use inductive methods to personalize their behavior, and we report some ongoing research that extends these ideas in the automobile environment.