Logo der Universität Wien


Modelling Method Evolution

A modelling method, as a software artifact, as well as a conceptual artifact, is constantly evolving. It is extended, trimmed, changed, split into multiple methods, or merged together with another method. These changes have to be tracked for various reasons: (1) to avoid reimplementation of existing functionality, (2) for future reuse, (3) statistics, and (4) visual representation. The tracking of changes is more complex than source code management in software development, because we are not working with source code, but with interconnected modelling method constructs: classes, relations, model types, notations, algorithms and mechanisms. Some of these construct are saved as scripts (e.g., notations, algorithms and mechanisms), some as objects in a repository (e.g., classes, relations, model types). Task at hand is to research the key requirements for managing modelling method evolution, and to implement a technical solution that is able to tackle with these challenges.

Design and implementation of a semi-automated business process composition mechanism

Business process composition is a key requirement in modern virtual enterprises, where previously isolated processes are gathered together under common process ownership and redesigned in order to improve the overall process performance. However, there is a significant challenge in supporting the new process owner with building the master process in a fashion that minimizes his effort and input. This topic proposes the investigation of current process composition practices, followed by the design and implementation of an ADOxx mechanism for composing existing business process models according to the modeller preferences. It should also include the design of a mechanism for capturing the modeller preferences regarding this composition.

Text Extraction from Various Internet Sources

In today’s world amount of information found on the Internet is increasing exponentially. Without the tools like the internet search engine (e.g., Google, Bing, etc.) or RSS feed that filter the information, provide the relevant content and brief summary, it would be impossible to find the needed information. Task at hand is to conceptualize and implement an internet text crawler that can search specified webpages for the information according to the given parameters. Found information needs to be stored in a repository in a form that can be broadcasted (e.g., CNN bulletin). For example, a text crawler is configured to search the www.openmodels.at website (including the sub-websites) for the phrases: business engineering, evaluation chains, and enterprise. The extracted bulletins could be: “The Business Engineering Navigator (BEN) has been developed at the Institute of Information Management, University of St. Gallen, Switzerland.”, “The concept of Evaluation Chains has been developed to support the analysis of complex economic situations and been applied first to the as-sessment of enterprise modelling.”, and “At any stage of its development, the enterprise knowledge model can serve as the means of understanding and communication between the participants in the EKD process (as well as others wishing to get a view of the enterprise).”, which will be saved in a re-pository and broadcasted on OM-TV in a few minutes. Research challenges relevant to the integration of the text crawler and the broadcasting service (e.g., OM-TV) should also be discussed.

Implementation of a Web process stepper based on BP models

One of the key goals of Business Process Management is the assimilation of processes and their cor-responding execution-related best practices by employees, assuming that process redesign and optimization has been performed. In order to train or familiarize employees with the new processes and the resources involved in executing them, various tools can be built, following the logic of underlying business process models. This topic proposes the implementation of an interactive process stepper as an AJAX web application, taking as input the exported XML structure of an ADONIS business process model.

Multi-View Modeling in ADOxx

One of the currently discussed topics in the research field of modeling and modeling tool development is Multi-View Modeling (MVM) and the development of MVM tools. MVM is characterized by the provision of a set of interrelated views, altogether realizing the model of some system under study (SUS). As these views visualize different aspects that are correlated to each other and are based on the very same SUS, consistency between these views is vital for the efficient application of a multi-view modeling method.
Meta modelling platforms, have proven to be very efficient in the development of modeling tools for domain-specific modeling methods. However, the specifics of MVM are not regarded properly until now.
The goal of this master thesis is to investigate the specifics of multi-view modeling methods on a theoretical/conceptual level. Consequently, the suitability of the ADOxx meta modeling platform for MVM shall be discussed. The thesis shall include the implementation of an add-on for ADOxx, acting as a proof-of-concept (POC) for the postulated requirements. This add-on shall foster the efficient specification of multi-view modeling methods based on ADOxx.
For further information, please write an E-Mail to: domenik.bork@univie.ac.at

Software Engineering Approaches for the Development of Modeling Methods

The effective application of modeling methods considerably depends on the availability of modeling tools. In order to cope with the complexity pertinent in modeling methods and the specific requirements of modeling tools, e.g. intuitive usage, extensibility and effective model management, conventional software development approaches, e.g. the Waterfall Model, are not adequate.
The goal of this master thesis is to investigate existing (software or modeling tool) development approaches and discuss their suitability for the development of modeling tools based on meta modeling platforms. The result of the thesis should be a process, constituting several steps in a sequential way leading to a software tool for a modeling method. For each step, specific tasks, supporting tools/services, skills required, input and output etc. should be defined on a general and abstract level. This would enable the application of the process in the broad area of modeling tool development, i.e. using the OMiLAB and Open Models Initiative communities.
Identified services or tools that are helpful in a certain step of the process should be implemented using the functionalities provided by the ADOxx platform and using the facilities of the OMiLAB.
For further information, please write an E-Mail to: domenik.bork@univie.ac.at

International PhD Proposal

Title: Enterprise Engineering Environment to support Risk Management in SME Servitization projects

Location: University of Vienna, Austria, Research Group Knowledge Engineering, with some periods in France.
International collaboration: Ecole Nationale Supérieure des mines de Saint Etienne (ENSMSE), France, Institut FAYOL.
Scientific field: Computer Science, Business Informatics

See more information (PDF-File)



Bei Fragen wenden Sie sich an den Betreuer o.Univ.-Prof. Prof. h.c. Dr. Dimitris Karagiannis (dk@dke.univie.ac.at oder DW 78910)

Universität Wien | Universitätsring 1 | 1010 Wien | T +43-1-4277-0