AbstractSoftware engineering is the practice of conceiving and developing tools to support business procedures. While it has made business more efficient, it has also crystallized new challenges. One of the major challenges in software engineering is to develop adequate software that effectively tackles business requirements. Software engineers are in charge of designing tools appropriate to the stakeholders’ needs, and they expect from the latter that they express their requirements as clearly as possible. Effective communication among those people is essential.
The quest for appropriate software brought out several software engineering methodologies, among which the model-driven engineering. Model-driven engineering is a branch of software engineering where the model is the core concept around which the development of the software is centred. In a nutshell, a model is a simplification of the domain of interest and this simplification allows to focus on the chosen concerns without being overloaded with irrelevant pieces of information.
Model-driven engineering, and more generally software engineering, have not dedicated so much research effort to the challenge of effective communication with stakeholders. Languages’ advances have largely been centred on defining a precise semantics and a well-formed abstract syntax. These efforts were necessary as the latter are pre-requisites to ensure that the applications coded according to the language behave as expected. However, they do not address the challenges related to communication among stakeholders.
In this dissertation, we investigate how SE models can be improved in order to ease the communication between software engineers and business stakeholders. Most of the software engineering models are represented as graphs, with nodes and edges. Our work focusses on making these nodes semantically transparent, that is to ensure that their appearance suggests their meaning. To achieve this objective, we adapt the Web 2.0 philosophy by empowering the end-users of the notation in the symbol design process. More precisely, we define a method consisting of a series of empirical studies to get end-users design and evaluate the most semantically transparent symbols for the i* goal modelling language. Hence, we provide not only such a structured method, but also a new set of semantically transparent symbols for i*.
|Date of Award||14 Oct 2016|
|Supervisor||Patrick HEYMANS (Supervisor), LAURENT SCHUMACHER (President), Bruno Dumas (Jury), Jean-Luc HAINAUT (Jury), Jean Vanderdonckt (Jury) & Daniel Amyot (Jury)|
- information visualisation
- visual symbol
- cognitive effectiveness
- visual notation
Attachment to an Research Institute in UNAMUR