Probabilistic Formal Modelling to Uncover and Interpret Interaction Styles
This work addresses the problem of interpreting user behavior for app designers, but it is incremental as it applies existing methods in a novel combination to a specific domain.
The study tackled the problem of uncovering user interaction styles in a mobile app using unsupervised computational methods, resulting in the identification of distinct styles that changed over time, with clear differences between initial usage (first day/week/month) and later periods (second and third months).
We present a study using new computational methods, based on a novel combination of machine learning for inferring admixture hidden Markov models and probabilistic model checking, to uncover interaction styles in a mobile app. These styles are then used to inform a redesign, which is implemented, deployed, and then analysed using the same methods. The data sets are logged user traces, collected over two six-month deployments of each version, involving thousands of users and segmented into different time intervals. The methods do not assume tasks or absolute metrics such as measures of engagement, but uncover the styles through unsupervised inference of clusters and analysis with probabilistic temporal logic. For both versions there was a clear distinction between the styles adopted by users during the first day/week/month of usage, and during the second and third months, a result we had not anticipated.