author: Zarimir Mitev
title: Measures for Interaction Complexity
keywords: measures
topics: Algorithms and Data Structures , Software Technology
committee: Arend Rensink
started: May 2019
end: July 2019



One of the main topics in software development is concerned with measuring the complexity of interactions. In this paper, the focus is on user-machine interactions. Current approaches for estimating user perception of an interaction before it is fully developed are expensive in terms of time, money and effort. Furthermore, when it comes to data collected for measuring the complexity of a given interaction, there is currently not an established methodology of reusing such data for measuring another interaction. In this research a new methodology is proposed which could make it possible to systematically collect data from users about few primitive interactive actions individually and then reuse the same data for the estimation of the complexity of arbitrary interactions composed of such actions. The validity of the proposed methodology is measured based on a sample of forty-three randomly selected students from the University of Twente and it is then compared with the average measured user perception accuracy.