Name: RAMILTON COSTA GOMES JÚNIOR
Publication date: 14/08/2015
Examining board:
Name |
Role |
|---|---|
| MONALESSA PERINI BARCELLOS | Internal Examiner * |
Summary: The i* framework is a Requirements Engineering language. Today, the community developing i* is relatively big. These developers, who are geographically dispersed, tend to attribute different meanings to i* constructs of the language. One may argue that due to the social intention behind i* modeling, a certain degree of freedom is convenient and these slight changes
should be acceptable. But on the other hand, it is our belief that this represents a barrier in terms of promoting the framework, creating serious problems, such as: a) difficulty in efficient communication among the communitys experts; b) increase in the languages learning curve by novices; and c) lack of acceptance by industry; and d) syntactic interoperability and existing semantics in various dialects. In the past few years, the community became aware of this problem and several attempts have been made to create a uniform use of the i* language. Although we recognize there are significant results in this direction, these attempts are not successful to solve the aforementioned problems, simply because these approaches are purely syntactical, not targeting the semantics behind the languages concepts. Going beyond syntactic questions, since 2006, researchers have been involved in an effort to define a common ontology to provide the semantics to the core concepts of i*. As a result of this approach, it is possible to provide a series
of modeling guidelines, here named ontological guidelines, to support the modeler in the use of the i* constructs. In this dissertation, we present an empirical study created to validate the ontological guidelines. For that, we propose an experiment made in a controlled environment, in which i* models are completed by two groups: one using the ontological guidelines and the other that does not have any contact with such guidelines. Results show that for more experienced conceptual modelers, the guidelines effectively represent a gain in providing higher quality models. For beginners in conceptual modelers, however, results are not equally promising. Based on the results of the experiments, this dissertation also proposes the creation of a plugin that
supports modelers beginners in building i * models compatible with the ontological guidelines, so that, little by little, the modeler learn and become more autonomous in the use of such guidelines.
This support occurs in the interaction through a dialogue between plugin and modeler, causing the ontological guidelines are useful in the practice of building models i *
