AbstractVisual notations are commonly used in Information Systems (IS) to represent information by using a structured set of rules and symbols. Nowadays, visual notations are used in almost every field of IS. Over the years, and since the first modelling language of Von Neumann, visual notations drew the attention of most research in IS. However, the research efforts are focusing more on semantic issues than on visual representation issues. Indeed, most of nowadays’ modelling languages do not provide any design rationale about their visual syntax, i.e., the symbols used in the visual notation. In general, modellers rely only on common sense to design the visual syntax of notations. To analyse and design notations, Moody’s Physics of Notations (PoN) relies on scientific foundations from various fields such as Psychophysics, Cognitive Psychology, and Graphic Design. However, an operationalization of the PoN theory remains crucial. Indeed, without any operationalization, a theory would appear to be limited to be applied on concrete visual languages.
Moreover, it appeared that the PoN was used differently from one researcher to another. This is due to the ambiguity caused by the need to refer to self-interpretation (i.e., intuition) on specific points of the PoN. An operationalization would reduce such an ambiguity. These reasons lead us to propose an operationalization of the PoN. To this end, this thesis will define a set of metrics to support two of the PoN principles. This thesis draws upon Störrle and Fish’s attempt to operationalize the PoN and on various work of Genon et al. who applied the PoN to analyse the cognitive effectiveness of modelling languages (i.e., optimisation of how visual notations are processed by the human mind). This operationalization of the PoN leads us to examine the degree of scientificity of the PoN by using a meta-theory (i.e., a theory to analyse and classify other theories). This theoretical analysis reveals the PoN principles that need to be more developed in future research. Finally, we propose to apply our operationalization on a modelling language developed by Mussbacher: the Aspect-oriented User Requirements Notation (AoURN). AoURN extends ITU-T’s User Requirements Notation (URN) to support encapsulation of crosscutting concerns in requirements models.
Furthermore, AoURN is the only standards-based and graphical approach to bring aspect-oriented modelling (AOM) to a requirements language. Moreover, the application of the PoN to AoURN is alone a valuable contribution.
|Date of Award||5 Sep 2014|
|Supervisor||Patrick HEYMANS (Supervisor) & NICOLAS GENON (Co-Supervisor)|