Keywords: Machine-processable Business Documents, Service-oriented Systems, Product Line Engineering, Evolution, Co-Evolution, Standardization, Variants, Variability
Astract: Standardized business documents specifying the structure and the semantics of information are essential for enabling information exchange in electronic commerce. The United Nations Centre for Trade Facilitation and Electronic Business (UN/CEFACT) provides a conceptual approach named Core Components (CCs) for defining business documents. In particular, business documents are defined using conceptual models serving as a basis for generating transfer syntax artifacts, such as XML Schema or RelaxNG schemas. These artifacts are used for defining service interfaces in systems, implemented following the Service-Oriented Architecture paradigm. Different subsidiaries of UN/CEFACT's International Trade and Business Process Groups (TBGs) are involved in defining such reusable building blocks. For instance, TGB3 represents the transport and logistics domain and TGB18 is active in the agriculture area. Consequently, the resulting standardized business documents are defined in an all-embracing manner covering the requirements of various business domains. Therefore, for the application in concrete business scenarios, standardized business documents need customizing for fitting a particular business context.
This results in a subset of the standardized business document which we consider as variants of the standardized business document. However, the CCs approach does not provide concepts for managing variants of business documents. Furthermore, the highly dynamic environment of electronic commerce with ever-changing market requirements typically impacts the structure and the semantics of information conveyed in electronic business transactions. However, evolution of standardized business documents has received little attention so far. For addressing these shortcomings we present the following contributions in this thesis. First, though the domain of business document standardization is not considered as a traditional domain in Software Engineering, we argue that concepts from Product Line Engineering (PLE) support us in dealing with business document variants. Thereby, we tackle the challenges from a different perspective that has not been considered so far. Second, we analyze and formalize variability in the CCs approach by employing, adapting, and extending existing variability concepts from PLE. Third, variability mechanisms are exploited for supporting the process of customizing standardized business documents.
Fourth, concepts are developed for managing access control in the context of PLE for business documents. Fifth, we define a reference model for the collaborative development and maintenance of the reusable building blocks which represents one source of evolution in the CCs approach. Sixth, we present evolution patterns for making evolution challenges more tangible. Seventh, we define model management operators for dealing with evolving business documents. Eighth, the concepts of Service Interface Adapters and Interoperability as a Service are introduced for translating between different versions of a particular business document.

Keywords:
Astract: Electronic data interchange (EDI) is the domain of exchanging business documents in a structured format. Nowadays, these systems take more and more advantage of service-oriented solutions. Nevertheless, a rigorous approach to customize the input and output messages of these services to the context of a business partnership is required. In this paper we suggest a model-driven approach to develop the XML Schemas of input and output messages of Web Services. Since classical data modeling approaches like regular UML class diagrams are not sufficient for modeling business documents, we base our approach on the UN/CEFACT Core Components Technical Specification (CCTS), which does not come with a presentation syntax. Thus, we deliver a UML profile extending class diagrams by CCTS concepts, called UML Profile for Core Components (UPCC). UPCC models are transferred to equivalent XML Schemas following well-defined naming and design rules. In order to allow for an easy search and retrieval of core component business document definitions, a registry is needed. The registry has to handle UPCC models - that are exchanged and stored in XMI - as well as their XML Schema equivalences. In this paper, we extend the ebRIM registry meta model for the special purpose of registering core component artifacts and defining their inter-dependencies.

Keywords:

Keywords:
Astract: Integrating Semantic Web concepts into the domain of e-business is a hot topic. However, most of the efforts spent so far concentrated on the improvement on B2C (business-to-consumer) e-commerce applications, achieved by semantic enrichment of information. With the growing importance of Service Oriented Architectures (SOA) companies started to move into the section of the Electronic Data Interchange (EDI), where applications exchange their business information semi-automatically. This B2B (business-to-business) electronic commerce is driven by aligning the internal business processes of companies to publicly available business processes. Thereby companies often do not consider the economic drivers of their business processes, which leads to incompatibilities between management, administration and technical layers. This chapter covers the two major domains of e-business / e-commerce, namely B2B and B2C. In the first, a model-driven approach towards B2B IT solutions is introduced, covering semantic aspects dealing with business models, business process models, and business document models. In the second application domain, the basic concepts of Semantic Web in the area of B2C electronic commerce are examined using a representative example from the e-tourism domain.

Keywords:
Astract: Conceptual languages to describe business models - not business process models - become increasingly important. One of the most expressive approaches towards business models is the Resource-Event-Agent (REA) ontology. REA has a substantial back-up in economic and accounting theory, but did lack a dedicated representation language for many years. Recently, we have introduced a domain specific modeling language including a REA meta model and a pertinent graphical notation. In this paper we build upon our previous work, but aim for a representation language that is able to serve as a serialization syntax also for other REA modeling approaches using UML class diagrams or OWL, etc. As a result we come up with REA-XML, an XML schema that allows a precise, tool-independent representation of REA models and that may also serve as an REA model exchange language between different tools.