Transformation of Algebraic Specifications Into Ontological Semantic Descriptions of Web Services

Abstract

The accurate description of service semantics plays a crucial role in service discovery, composition and interaction. Most work in this area has been focused on ontological descriptions, which are searchable and machineunderstandable. However, they do not define service functionality in a verifiable and testable manner. In contrast, formal specification techniques, having evolved over the past 30 years, can define semantics in such a manner, but they have not yet been widely applied to service computing because the specifications produced are not searchable. There is a huge gap between these two methods of semantics description. This paper bridges the gap by advancing a transformation technique. It specifies services formally in an algebraic specification language, and then, extractsan ontological description of domain knowledge and service semantics as profiles in an ontology description language such as OWL-S. This brings the desired searchability benefits. The paper presents a prototype tool for performing this transformation and reports a case study to demonstrate the feasibility of our approach.

Monic Testing of Web Services Based on Algebraic Specifications

“…The theoretical mathematical foundations have expanded from initial and final algebras to behavioural algebras [19] and co-algebras [20], [21]. Both functional and behavioural properties can thereby be specified and it seems that this approach may be suitable for web services [5], [22].…”

Quality model and metrics of ontology for semantic descriptions of web services

“…(2) Ontology Generator, which takes the parse trees of a collection of specification units as input, and generates domain ontologies and ontological descriptions of web services according to the rules given in Ref. [18]. Figure 6 shows the interface for this component.…”

Automated Testing of Web Services Based on Algebraic Specifications

“…• independent of implementation detail, which is appropriate, because no such detail will be available about third-party services, • highly modular in a manner that suits flexible composition, as is also required by service-oriented engineering, • easily translatable into ontology-based semantic descriptions, facilitating registration and then dynamic discovery and support [20], [21], • suitable for specifying dynamic behaviour too, when extended to co-algebraic specifications, • written in a notation that is easy to learn and to understand, according to empirical studies such as [22]- [24]. Moreover, as we shall see, algebraic specifications allow the whole testing process to be automated, including test case generation, test execution and test oracles to determine the correctness of test results.…”

“…There are two ways a unit can be constructed from another: extension and usage. As in [20], [21], we assume that the specification of a software system is wellstructured in the following sense.…”

“…Let (Sp, Σ) be a given system signature and s ∈ S be any given sort. We defined the notion of valid terms in [20], [21] that can be used in the specification unit of sort s as s-terms. Each s-terms is also typed and its type is w ∈ W s .…”