Keywords: Procedural modeling, traditional Balinese architecture, intangible heritage, photogrammetry, parametric modeling
Astract: raditional Balinese architecture is based on ancient Bali-Hindu philosophy which can be observed in traditional built environments on Bali. The underlying architectural principles are very complex containing intrinsic “parametric rules” based on both a philosophical and a building tradition. The parameters are in relation to the religious belief system as well as being derived from the bodily dimensions of the “head” of a building. We aim at determining in which way these underlying building principles can be preserved by contemporary digital means. As a hypothesis we test a parametric formalization as CGA shape grammar made operative using the CityEngine. In order to formalize these building principles the following methods were devised: identification and collection of appropriate literature in English and Indonesian, partial translation of available material to English, field trips for contextual information and additional knowledge, and transformation of all sources of information into a shape grammar. An interactive visualization of ancient Balinese building principles is exemplified through the traditional house compound, the “rumah”, demonstrating the strength of a parametric approach. Moreover, it includes both, tangible and intangible heritage aspects by analyzing the appearance of (tangible) built heritage, as well as the (intangible) underlying philosophy and traditional knowledge. Our contribution aims at supporting the understanding and preservation of the architectural heritage of Balinese building principles.

Keywords:
Astract: Enterprise Resource Planning (ERP) Systems encompass the administration of all information about resources that are needed for companies to run their business. They support several functional areas, like accounting, manufacturing, and sales in form of modules that are integrated by a single database where all business relevant information is stored. In order to guarantee a flawless and productive use of the system, the economic phenomena underlying the business of the company have to be reflected in the user interface as well as the data structure itself. Usually ERP systems are purchased by customers from vendors in form of standard software. Such a software supports a predefined set of functionality and has to be further customized to the specific needs of an enterprise, which is not a trivial task and often leads to additional costs. Furthermore standard software only supports specific processes that are based on best-practice assumptions of the vendors. Therefore adjusting ERP systems to changed market demands is hard since they are missing a business semantic base. Changed business needs can often only be represented by drastic changes in the data structure or the code which can lead to inconsistencies. The REAlist project uses a model-driven approach to overcome the aforementioned problems with existing ERP systems and enhance their adaptability. Business needs can be represented in an easy and unambiguous way as business models. These models define the features and look of the ERP system. REAlist uses the Resource-Event-Agent (REA) ontology as a business modeling language since it was initially proposed to support the implementation of IT-Systems and is related to data modeling. REA allows the specification of events that have happened or are happening in the near future, resources affected by the events, and agents participating in the events. Furthermore, policies and commitments can be defined which are both important concepts of ERP systems. The underlying database of the REAlist project (REA DB) is based on REA and holds the business models as well as the business data. Its main part is generic, meaning that changed business needs can be applied to the ERP system without changing the data structure. Instead of using the classic class-diagram-like representation of REA, a domain specific language (REA-DSL) is used to simplify the creation of REA business models. The aim of this thesis is to undertake the first steps of the REAlist project and create a mapping from REA-DSL business models to the REA DB. Furthermore user interfaces are automatically generated based on the saved models during runtime to reduce the effort that is needed for the customization tasks in existing ERP systems.

Keywords:

Keywords:
Astract: Special Issue on A Roadmap for Business Informatics: In this special issue of the EMISA journal we include seven papers, each based on a IEEE CBI 2013 keynote introducing a research domain in Business Informatics. Evidently, these papers are neither classical research papers nor pure surveys, since they focus to a large extent on the "future", i.e. the open research challenges (without providing a solution). In the following, we define the scope of the seven research domains and in parentheses we name the author(s) who introduce( s) the domain by a paper presented in this special issue.

Keywords: components modeling
Astract: Business document standards usually cover a hierarchical structure of thousands of elements that may be relevant in any business context (any geopolitical region, any industry, etc.). In order to use a business document standard in a specific context, user groups define so-called business document implementation guidelines based on a smaller subset consisting usually of 3 - 5% of the overall elements. When one defines a new implementation guideline for a specific business context, one has always to start from scratch, which is time-consuming and also leads to heterogeneous interpretations of the standard. It is our goal to speed up the development process and to create more homogeneous implementation guidelines by learning from existing models. If we could assign a formal context to existing implementation guidelines, we may predict the subset of a new implementation guideline for a given context. We especially consider implementation guidelines built on the top of the Core Components Technical Specification (CCTS). These guidelines consist of business context specific data building blocks which are restricted from more general, semantically interoperable Core Components. In order to share, search, and (partially) re-use context specific restrictions of Core Components it is essential not only to store the restrictions, but also a business context model where these restrictions are valid. Therefore, we develop the Enhanced UN/CEFACT Business Context Model (E-UCM) and the Business Context Ontology Model (BCOnt) for representing business context in the domain of Core Components. This thesis proposes an approach to contextualize already existing Core Components by means of our business context models. This contextual information is also used to predict a subset for to-be-developed electronic business document implementation guidelines based on existing ones. The underlying algorithms calculate degrees of business context match, detect different types of mappings between existing Core Components and generate Core Component based contents of new implementation guidelines. Our research has been conducted following the design science research process. The corresponding evaluation is interpreted as a build and evaluate loop iterated a number of times before the final approach was developed. We evaluate the business context models on the basis of 16 evaluation criteria. The feasibility of the business context aware Core Components modeling approach is demonstrated by a prototype implementation. The analysis of the calculated precision and recall rates proves that our approach holds not only in theory, but also in practice.

Keywords:
Astract: With the introduction of OMG's Foundational UML (fUML) standard, that precisely defines the execution semantics for a subset of UML, and the conforming virtual machine, completely executable systems can be built and executed with UML. However, the full potential of having executable models has yet to be unleashed. An important aspect that increases the potential of executable models is the ability to re-use existing software components since that reportedly increases the overall quality and productivity of the software development process. Furthermore, large-scale software that is produced nowadays, involves the usage of a high number of existing software components primarily in form of software libraries (i.e., APIs provided for the used programming language). This thesis identified that the fUML standard does not offer a procedure to use software libraries. In fact, creating models that build on top of software libraries is not foreseen in the fUML standard. On the contrary, the standard foresees its extendability through the Foundational Model Library. Yet, doing so requires implementing model libraries that basically mimic the functionality provided by currently existing software libraries. This approach imposes a significant drawback. It requires a huge amount of dedicated joint effort to build a set of libraries having similar functional coverage and sophistication as the existing set of software libraries. The research question of this thesis is as follows. Is the fUML virtual machine extendable, such that it allows the execution of models referencing external software libraries? Within this work, an approach has been elaborated that enables to use external software libraries in fUML models. The applicability of this approach has been shown by implementing a prototypical Integration Layer that is able to integrate Java libraries with the fUML virtual machine such that the modeler can benefit from the advanced and complex functionalities provided by those libraries. This prototype focuses on creating instances of library classes and calling library operations. Moreover, a two-step procedure to make existing libraries available for their usage in fUML models, has been implemented. While conducting several case studies, experiences have been gained that led to further enhancements of the prototype and to the following conclusion. The fUML virtual machine can be extended, such that it allows the execution of models referencing external libraries. Nevertheless, to broaden the applicability of the implemented prototype, and therefore increase the scope of applicable modeling scenarios, an in-depth investigation on common library use cases and their following implementation is recommended.

Keywords:
Astract: Today, modeling and programming constitute separate activities carried out using modeling respectively programming languages, which are neither well integrated with each other nor have a one-to-one correspondence. As a consequence, platform and implementation details, such as the usage of existing software components and libraries, are usually introduced on code level only. This impedes accurate model-level analyses that take platform-specific decisions into account as well as the direct deployment of executable models on the target platform. In this work we present an approach for integrating existing software libraries with fUML models-an executable variant of UML models for which a standardized virtual machine exists-not only at design time but also at runtime. As a result of that, the modeler is empowered with the capabilities provided by existing software libraries on model level. Our approach is evaluated based on unit tests and initial case studies available in the ReMoDD repository that assess the correctness, performance, and completeness of our implementation.

Keywords: model-driven engineering, verification, model checking
Astract: The aim of this master thesis is to reduce the conceptual gap between software modeling and model checking. While model checking is successfully applied for hardware verification, it is not widespreadly used in model-driven engineering (MDE). Thus, we tried to reduce this gap by combining modeling and model checking concepts. This thesis first describes the history and basic idea of both MDE and model checking with a focus on the technologies used in this thesis. Before presenting our new approach, existing solutions are compared. Most approaches propose to extend the Object Constraint Language (OCL) by temporal aspects. This allows to describe the behavior of a software system additionally to various properties of static models. However, one of the main missing features in general seems to be a user-friendly representation of the verification result helpful for debugging. Often, the technical spaces are changed. With our solution we provide (i) a temporal OCL extension based on the Computational Tree Logic (CTL) and (ii) an OCL extension that introduces path selectors to extract interesting system configurations from the state space. Both OCL extensions were formally defined and implemented. We describe systems in terms of state spaces consisting of EMOF-model states and state transitions containing a mapping between model elements of different states. The system behavior is specified using an initial Ecore model and graph transformations based on the Henshin tool2. The approach, however, is designed to be flexible enough to allow an easy integration of any kind of behavior specification as long as a suitable state space can be derived thereof. Our model checking framework is developed with a focus on delivering not only the results, but also making the system behavior leading to the result comprehensible by providing a suitable tool including a web interface. The implementation was evaluated in terms of performance to find out the maximum evaluable model size and query complexity. Further, a qualitative user study was conducted for evaluating the CTL extension and the tool. The results of this study indicate that both the CTL-based extension of OCL as well as the tool are a promising first step to integrate model checking in the MDE life cycle.

Keywords:

Keywords:
Astract: The evaluation of an organization's performance may also consider the assessment of inter-organizational relationships (IORs). However, the evaluation of IORs is typically based on success factors, such as trust, which are difficult to be measured quantitatively. In this paper, we present a framework supporting inter-organizational performance evaluation which integrates (i) a bottom-up approach supporting the identification of Key Performance Indicators (KPIs) from business information, event logs, as well as process models, and (ii) a top-down approach for measuring business performance on the strategic level based on the Balanced Scorecard (BSC) method. In order to prove the feasibility of the framework, we present an inter-organizational performance analysis case study of a beverage manufacturing company. The case study shows that the framework enables (i) the derivation of quantifiable KPIs from operational data and (ii) the alignment of KPIs with business objectives allowing an evaluation of IORs on the strategic level.