Anthony A. Hock-koon

Defining Metrics for Loose Coupling Evaluation in Service Composition

Service Oriented Architectures (SOA) mainly rely on the service composition mechanism which supports the reusability. Numerous researches focus on this thematic in order to satisfy all the objectives of quality fixed by the SOA community. One of these objectives is the notion of loose coupling between the compositions services. However this notion is intuitively understood and tangible, its definition lacks formalism. This unclear picture of the loose coupling is reflected in limited metrics which do not allow for its quantitative and objective evaluation. In this paper, we propose a set of different metrics which are associated with a clear definition of the loose coupling notion. These metrics act as guidelines for architects to develop their composite service. They are gathered in a formula which allows for evaluating the global coupling of a composition of services. Moreover, they are also used as comparison criteria for classifying existing composition models.

Composite service metamodel and auto composition

Service Oriented Architecture (SOA) is a computing paradigm which promises to provide transparency to resource access by exposing the available resources as services. SOA is based on service composition for application development or service development. Numerous researches focus on service composition. Each of them tries to solve some particular problems. This paper intends to reify the relevant notions of these models in a Composite Service MetaModel. This metamodel defines all interlaced features and provides a global and explicit vision of the service composition. Moreover, this approach allows the specification of the auto composition process: the composites ability to dynamically modify its architecture and its composition logics according to the environmental context.

The Product-Process-Quality Framework

In this paper, we present the Product-Process-Quality Framework which provides a clear understanding of the differences between Component-based software engineering (CBSE) and Service-oriented software engineering (SOSE). We intend to provide a conceptual comparison to architects who have to deal with these two software development paradigms. This work completes existing ones which focus on the distinction between Objects and Components.

Many to many service discovery: a first approach

Dynamic service discovery is one of the main concepts which define the Service Oriented Architectures (SOA). This mechanism ensures loosely coupled services. It supports the reusability of services and the flexibility of applications. This paper provides a new service discovery approach which enhances the number of services potentially discovered and thus, it multiplies the number of candidate services. This multiplication ensures a better selection of the most suitable services and more alternatives is case of failures.

Vers une meilleure compréhension de la différence théorique entre un composant et un service

The distinction between Objects and Components is well-known, but the distinction between Components and Services is much less clear. Therefore, we define the Product-ProcessQuality framework which intends to provide a clear understanding of the differences between Component-based software engineering (CBSE) and Service-Oriented software engineering (SOSE). Moreover, this framework handles the Object orientation (OO) to provide a global point of view of these three styles of software engineering. MOTS-CLES : Objet, Composant, Service, Cadre de Comparaison Conceptuel

Toward the Definition of the Loose Coupling Notion in a Composite Service

Service composition is one of the most interesting challenges facing the Service Oriented Architecture (SOA) paradigm. Its accomplishment promises to maximize the information reuse. Service composition mechanism wants to facilitate application development by combinations of existent resources which are exposed as services. This mechanism is associated with the notion of loose coupling. The diminution of dependencies between services is intuitively understood and gives immediate benefits. However, this intuitive understanding hides a lack of formal theories. In this paper, we intend to provide a clear definition of the loose coupling notion. We reuse the concepts introduced by our Composite Service MetaModel to identify its associated metrics. These metrics will allow for the comparison of existent service composition approaches. Moreover, they define some development guidelines for architects.

Specifying loose coupling from existing service composition approaches

The loose coupling notion associated with the service composition is a key concept which defines the Service Oriented Architectures (SOA) paradigm. Given that this notion is intuitively understood, its definition lacks formalism. Moreover, the existing evaluation metrics are limited and cannot take into account all the specificities of the SOAs composition mechanism. In this paper, we present a set of metrics based on a clear definition of the loose coupling. We combine these metrics in a formula which calculates a weight. This weight allows for clear measurements of a composites coupling.