A qos ontology-based component selection

TL;DR

This paper introduces a QoS ontology to facilitate semantic and effective selection of software components based on non-functional QoS requirements, addressing heterogeneity issues in component descriptions.

Contribution

It proposes a formal QoS ontology and a two-step selection process combining semantic matching and ranking for improved component selection.

Findings

Effective semantic matching of QoS descriptions.

Successful application in multimedia component domain.

Enhanced selection accuracy with ontology-based approach.

Abstract

In the component-based software development, the selection step is very important. It consists of searching and selecting appropriate software components from a set of candidate components in order to satisfy the developer-specific requirements. In the selection process, both functional and non-functional requirements are generally considered. In this paper, we focus only on the QoS, a subset of non-functional characteristics, in order to determine the best components for selection. The component selection based on the QoS is a hard task due to the QoS descriptions heterogeneity. Thus, we propose a QoS ontology which provides a formal, a common and an explicit description of the software components QoS. We use this ontology in order to semantically select relevant components based on the QoS specified by the developer. Our selection process is performed in two steps: (1) a QoS matching process that uses the relations between QoS concepts to pre-select candidate components. Each candidate component is matched against the developer's request and (2) a component ranking process that uses the QoS values to determine the best components for selection from the pre-selected components. The algorithms of QoS matching and component ranking are then presented and experimented in the domain of multimedia components.