Reconciling context-oriented programming and feature modeling

Details

Supervisors

Mens, Kim

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques

Abstract

The advent of connected devices, such as smart-phones, enabled the con- ception of several of software applications that are yet more intelligent. At the moment, the information in these applications is mainly encoded by individuals, but we can imagine in the near future that sensors may produce and provide it without any human intervention. With nowadays tools, it is very complicated to design smart software applications that are aware of their surrounding environment: program- mers must handle all existing combinations of the contextual information at once, which increases the complexity of the program exponentially with the amount of information that is considered by these applications. Some software engineering domains aim to ease the development of such applications, such as Feature-Oriented Software Development and Context-Oriented Programming. Although they have been studied in isolation and use different approaches, they seem to solve similar problems. Still, few work has been performed to reconcile ideas from both these domains. The goal of this thesis is to reconcile these two domains, in order to ease the development of applications with highly dynamic behavioural changes. To do so, a new formalism is introduced, so as to design a framework to build these applications. The framework is tested with a self-assembled context-aware application, an Emergency Response System that helps users in case of emergency.