A pattern approach for identification of opportunities for personalisation and automation of user interactions for the IoT

This paper introduces the 'Personal Spark' pattern which can be used to identify relevant fragments of human-human interaction as an initial step to inform requirements for adaptive user interfaces and data processing, in this case for user interactions of consumer Internet of Things.

This pattern is aimed primarily at helping user experience designers, in modern, iterative environments, with the challenges and opportunities of using data and Machine Learning as design resources, since the resulting enhanced interaction concepts often represent ideal use cases for Machine Learning.

The approach was tested with 23 User Experience designers who, overall, rated it as a very positive experience. This research reflects on a novel route to assist the conceptual phase of the user interaction design of consumer IoT. This is done by providing user experience designers with a pattern approach to help reducing the gap between the opportunities in human-computer interaction research and the practice of user experience design in the industry.