Towards a Framework for Empirical Measurement of Conceptualization Qualities

Abstract

Conceptualization development is central in modeling language design. As one of their first design steps, language designers need to decide on a set of concepts on which the language will be based and which can be understood and used by a population of modelers for characterizing and representing relevant domain information. Thus, exposing candidate concept sets to future users may offer insights on how well the concepts of choice are understood and distinguished from each other by those who will be called to actually use the language. We propose an empirical measurement framework to allow just that. The framework consists of an instrumentation approach whereby participants sampled from the user population classify domain expressions to the corresponding concepts, and a set of measurement constructs for translating participant observed data into design insights. A small case study is conducted to explore the feasibility and limitations of the proposed approach.