Community centric service methodology à [home]

Definition of “service” and “web service”

Previous discussion à [48]

A proposed service methodology is designed to transform how computer based information structure is used in everyday activities. We move in two directions,

(1) towards ontology mediated automated web services and

(2) towards the measurement and responsiveness of formative information structure.

The first direction is taken by a number of groups, example DERI; the second direction is taken only by a few, example Acappella Software.

Dieter Fensel is one of the leading figures at the Digital Enterprise Research Institute (DERI) in Germany and has developed, along with his colleagues a number of OASIS standards related to the use of information models, process models for web service discovery and SOA blueprints for contractual agreements about how business centric web services are to be made interoperable.

“Various proposals for automating the discovery of Web services are available. These often regard the terms service and Web service as synonymous (e.g. [Paolucci et al 2003], [Li and Horrocks, 2003]). We believe these two terms are not equivalent and, furthermore, it is relevant to distinguish them and to explore their relation, as done in [Priest, 2004] and [Baida et al, 2004]. We agree that this fundamental distinction is necessary in order to achieve scalable and realistic Semantic Web service Discovery.”

Fensel Dieter et al, “WWW or What is Wrong with Web Service Discovery” (June 2005).

This service methodology is being developed from a very broad definition of what a service is, and relates our definition of “service” to the concept of a transaction space. In this sense, we accept the position that Dr Fensel takes in the above quote. Our position takes an additional step, however; asking that the “service” be also very much of the type we see and experience as everyday transactions occurring between average people.

We would like “service” to mean even those that take the form of knowledge exchanges involving no economic trace. We include in our broad definition of “service” the exchanges in everyday discussions. We include “social service”. In this sense, not only are casual questions considered a request for a service, i.e., a tacit request to receive an answer; but also the act of discourse itself is both requiring and giving service.

“The need for such distinctions becomes more evident if we think about a prototypical e-business scenario: a user wants to get a specific service that provides some real-value for him. Web services are technological means for accessing or specifying services offered by some provider. In a sense a Web service is an access point to services of a particular provider. Users are not specifically interested in Web services but rather in the services that can be delivered by a specific provider. Clients think in terms of services whereas providers advertise sets of services they are able to offer to their clients (using a technical entity Web service).

Fensel Dieter et al, “WWW or What is Wrong with Web Service Discovery” (June 2005).

“Web services” must exist as well specified data flow if web services are to exist at all. So the sense of business pragmatics forces what appears to be an economic definition of service. The business community sees a service as “a business activity that often results in intangible outcomes or benefits” [Baida et al, 2004]. For the business community this definition is the be all and end all definition.

However, so much of what is valuable is not and (one expects cannot be) completely reduced to business activities. The core issue has to do with the difference between pragmatics and semantics [de Moor, 2005].

Specifically, we hold that the induction of new information structure in the moment during a period of crisis involves pragmatics. One has to be careful here. Our definition of pragmatics is grounded in the notion that new natural categories arise in transaction spaces, that as these new natural categories are arising there is a period of time in which the meaning is not well specified and cannot be given a precise, or formal, definition. Thus there cannot, always, be a complete formalization of the meaning of events while the events are occurring. Clearly the relative importance of pragmatics in each situation varies.

Pragmatics is by definition about the uniqueness of the situation. The full notion of “meaning” has not gone through a categorization and reification process, and thus cannot have been converted into formal semantics. The distinction between pragmatics and semantics parallels the distinction that Dieter Fensel makes about “services” and “web services”. Web services have formal notions of pre-established meanings, whereas “services” have a variable, and situational, pragmatic aspect that may not yet completely reduced to category. In some cases, the event structure and information about the event cannot be given precise semantic interpretation until later on.

If the driver of service requests is a decision intensive process [Bromberg et al, 2006] the delay between request for service and the natural establishment of categorical meaning may be the difference between life and death. In Part Two, we will return to Bromberg’s definition of decision intensive process, when we develop a framework based on a seven step SOA methodology [Cyrille, 2006] and Bromberg’s six characteristics of decision intensive process.

Our new proposal for an OASIS Community Centric Service Methodology TC is founded on these types of distinctions. One sees a service as the primary type of transaction occurring in any type of transaction space. For example a metabolic process might provide services to another metabolic process. A more relevant example may be a group discussion, about how to respond to a crisis. Such a group discussion might provide a social service to a President or Governor.

To be clear, we suggest that in all cases there are significant pragmatic axes where natural category is not fully determined by real event structure. In cases where organizations or individuals are struggling with a crisis, this pragmatic axis will provide the best origin of control, and so semantic formalism may even be harmful to optimal response. In Part Two we will make distinctions between event structure where services are being requested. In same cases we will have a powerful pragmatic axis and thus very little pre-arranged formal semantics. In other cases, the formalized interpretations of meaning will be sufficient for optimal automation of service requests and fulfillments.