Article Preview
Top1. Introduction
Many systems around us can be described by network models. The examples available are numerous and range from social networks, to the Internet, and to power grids. Understanding how services evolve as networks and the risks and gains of different topologies is also becoming increasingly critical for society (Spohrer & Maglio, 2010). The impact on society can be compared to contributions made in the areas of social networks and the Web. Considering every single individual on this planet as a potential owner of resources that are relevant for the provision of at least one real-world service, we would count seven billion service systems, and therefore information systems need to be in place to enable each individual to have an active role in service-centric societies.
The ability to construct service networks is the most basic requirement to understand the dynamics of global service-based economies, and their innovation. However, currently available techniques fall short of providing workable solutions as they are unable to deal with the automated description of open and rich relationships between services. In this article, we focus on the particularly challenging task of providing the foundations for constructing open semantic service networks (OSSN) by accessing, retrieving, and combining information from service systems and relationship models globally distributed. Service systems, relationships and service networks are said to be open when their models are transparently available and accessible by external entities and follow an open world assumption. Networks are said to be semantic since service systems and relationship models are expected to have a shared understanding regarding vocabularies, rules and semantic Web theories and technologies.
Current developments are targeting the computer-understandable description of services using comprehensive languages such as the Unified Service Description Language (*-USDL) (Cardoso, Barros, May, & Kylau, 2010). In the near future, these languages will allow formalizing the description of service systems in such a way that they can be used effectively for dynamic service outsourcing, efficient SaaS trading, and automatic service contract negotiation (when no ambiguity arises, we will simply use the term service to refer to a service system).
We take the challenge of developing a novel perspective on the global economy by connecting service models representing open business service systems (e.g. consulting, e-governmental, SaaS and educational services) typically provisioned by commercial companies and governmental agencies. The difficulties which we face in doing so differ significantly from those tackled by prior work in global distributed information integration and large scale systems. Compared to previous work in Service-Oriented Architecture (SOA), we target the study of business services which goes beyond the analysis of Web services (e.g. WSDL, SOAP, and REST) in complexity. Furthermore, SOA generally relies on top-down and BPM (Business Process Management) strategies to develop process models. We take a totally different approach: we follow a self-governing and bottom-up approach which generates network models by using relationships instead of using temporal and control-flow connectors.
This article is structured as follows. In Section 2, we provide a motivation scenario for the importance of constructing global service networks. Section 3 provides a literature review in this emerging field of science. Then, Section 4 suggests seven principles for community-based design of OSSN. Next, Section 5 presents an approach that can both maximize the usage of available service information and enable the location of related services in online marketplaces. Our solution involves four research activities: 1) identification of a suitable schema, language, or ontology to model services, 2) creation of an open and rich model to represent relationships between service models, 3) populate service and relationship models, and 4) development of an infrastructure to query and access distributed service models, and dynamically construct global service networks. Section 6 illustrates the value of OSSN in the context of several research projects. Finally, Section 7 provides the conclusion.