FixOnto: um método para enriquecimento semântico e verificação de modelos de características em LPS sensível ao contexto

Abstract

Dynamic Software Product Lines (DSPL), designed to manage the variability of self-adaptive systems at runtime, can be employed to systematize and maximize reuse in context-aware applications development. To this end, there are also Context-aware Software Product Lines (CASPLs) which are focused exclusively on supporting context-awareness. Feature models are the main way to represent similarities and variabilities in traditional, context aware and dynamic SPL. CASPL, focus of this work, contains, for example, Mobile and Context-aware Feature Model (MCFM), which includes the feature model and the context model. Although these models are the main representation of knowledge about a domain in SPL, they have limitations of expressiveness. For instance, relevant domain aspects (such as a UML artifact that is associated with a context) are not described in the MCFM. In addition, these models may contain inconsistencies that lead to the derivation of invalid products. To solve these two problems, the goal of this work is to propose a method to add semantics to the MCFM and perform automatic verification of the correctness and consistency of these models. A tool is implemented to evaluate the method and it is performed a demonstration of its use with a CASPL called Mobiline. As a result, it is observed the possibility of verifying that the models are correct, considering the implemented rules, and that the use of ontologies in the process of semantic enrichment allows the realization of semantic searches as well as traceability of features, contexts and artifacts.