On Applying Ontologies in Workflows : Ontology driven Workflows for increasing Flexibility and Quality and reducing Production Time and Cost

Abstract: The trend in business goes to higher flexibility of companies due to the manifold wishes of the customer, which challenges information systems. A consequence is the mass customisation, the effort to produce different variants of one product to meet the multifaceted needs of as many customers as possible. Supply-chains are not the only concept to take into consideration, but also value chains and value systems. The customer has to be willing to pay for a product, and every step in the value chain respectively value system has to add a value the customer is willing to pay for. For the companies to live up to these demands new structures have to be established, companies have to work in networks. Every company does what they do best, and are highly specialised.The focus in information systems has to be on the point where business partners communicate with each other, on a peer-to-peer basis, the interfaces. The transfer of information from one system to another can be a problem, especially when the system sending the information has other semantics than the receiver, leading to misunderstanding between the business partners. One possible solution for overcoming these problems is to explicitly consider ontologies – a shared agreement about the meaning of the concepts. The communication model shows how to ensure the transfer of content from one system to another, enabling business partners to transfer knowledge between each other, by using ontologies.This PhD thesis examines the effect and possibilities of the explicit consideration of ontologies in workflows, when used in a workflow management system. The investigation was done with help of a case with three units of analysis. The examination of the case was within an EU-project, having the goal to improve the quality and flexibility, as well as decreasing costs and time in companies in the automotive industry and their networks.A common process of all three units of analyses was the quotation process, which is presented in this PhD thesis. The first task was to investigate the ontologies of the companies, on a social level. The goal was to deliver the information necessary for the four ontologies, using amongst others the Socrates analysis.A second task was to evaluate the ontologies used in a prototype. The goal was to get the opinion of the future users about the prototype, in particular about the functions offered by the prototype using the investigated and formalised ontologies. The main goal within the evaluation was to compare the way of working and accomplishing results with and without the prototype. The prototype was evaluated, amongst others, according to the four dimensions quality, flexibility, cost and time.The measuring of effects by using the prototype of the workflow management system was one of the challenges of this PhD thesis. The data collected in the evaluation was categorised first according to the four phases of the global architecture, and second according to the four dimensions of the project. A matrix shows in what stage what dimension is occurring. The conclusion is that there are positive effects on the workflow when using the ontology-based prototype. The combined use of the workflow management system and the use of the developed ontologies showed clear and positive results on the four dimensions.Last but not least the PhD thesis provides suggestions for improvements when investigating ontologies and designing general ontology based systems. Two models, and how they interplay, are presented, first the 4-Level-Ontology-Model, and then the Intelligent Ontology Translator. The 4-Level-Ontology-Model provides suggestions how to classify ontologies of an organisation vertically, while the Intelligent Ontology Translator considers the transfer of content on a horizontal level. These two models, together with the global architecture, open up for a new architecture for communication between systems, the ontology driven architecture (ODA).