Direkt zum Inhalt
Eine algebraische Fläche vom Grad 6 (eine "Sextik"), die 65 Singularitäten besitzt.

Fields of research

The working group contributes solutions to the following software engineering challenges:

Developing mobile apps of high quality

Rapidly increasing numbers of applications and users make the development of mobile apps to one of the most promising fields in software engineering. Due to short time-to-market, differing platforms and insufficient adaption of existing test processes, quality assurance processes of mobile apps have not taken into account enough.

The project "Mobile Testing" (founded by the BMBF, in cooperation with the eXept Software AG) is dedicated to the development of a semi-automatic test process for mobile apps. Using this test process, mobile apps shall be tested concerning their requirements (specification-based test) as well as concerning more general quality criteria (specification-less test). In addition, the test process shall identify relevant context conditions of mobile apps and thereafter, it shall generate test variants for these contexts.

Continuous quality management of dynamic, partly uncertain research data

Data of a specific domain is often stored in several heterogeneous data sources and formats. These formats as well as the data itself may continuously evolve due to several reasons. On the one hand, knowledge may be ensured eliminating some uncertainties, on the other hand, new requirements may emerge as, e.g., to further integrate data. The quality of recorded data and data formats may vary a lot but a quality management (QM) process is often not pursued. 

The aim of the project KONDA (founded by the BMBF, in cooperation with the German Documentation Center for History of Art in Marburg and the Göttingen State and University Library) is the development of a continuous and adaptable QM process for research data on objects of material culture. This process shall be generic to be usable not only in this research domain, but also in various others. Furthermore, this process shall be continuous to ensure a high quality of data, data formats and transformations, all dynamically changing over time. Since quality management of uncertain data has been hardly investigated, we will focus on new modelling and quality assurance techniques especially for this kind of data. All newly developed techniques and processes shall be evaluated at the picture index of the German Documentation Center for History of Art in Marburg. This index contains over 1.7 million works of art and buildings and over 2.9 millions photographs. This evaluation shall show that they effectively improve this index.

Model driven engineering

Modern software systems tend to become more and more complex in structure and behavior. Model-driven engineering (MDE) lets us expect a better control of software complexity as applications are not developed purely by programming but also on a higher abstraction level using models. Instead of juggling with different implementation technologies, common development tasks are performed on domain specific models which allows to abstract from technology details. Model-driven engineering tools such as model extractors and code generators allow to bridge the gap between models and underlying program code. Exchanging underlying technology components does not have to lead to a new implementation of a software (component) but may be cushioned by new versions of MDE tools.

Domain-specific modelling. Model-driven engineering is usually based on models of domain-specific languages. For working with models, developers need integrated development environments (IDE) for these languages. Such an IDE shall consist of model editors and additional tools for effectively developing models including tasks such as simulation, transformation, version management, and quality assurance. Accessing a new domain, a new IDE for the corresponding modeling language has to be developed. 

The aim of our project "Meta modeling and graph grammars" (founded by the DFG, in cooperation with the University of Oldenburg) is to develop new concepts and tools for the model-driven development of IDEs for domain-specific modeling languages. Our intention is to make the development of such IDEs easier and more efficient. The new concepts shall be based on meta-modeling and graph transformation since domain-specific language can be adequately defined with these concepts. 

Model transformation.  Model transformation is a core concept of model-driven engineering. We distinguish exogenous model transformations on the one hand, translating or converting models from one language into another, and endogenous model transformations on the other hand, optimizing, simulating or, in general terms, changing models within one language. The project Henshin aims at providing a language and toolkit for the transformation of EMF-based models. Models of different projects (and potentially of different languages) may be synchronized using Triple Graph Grammars (TGGs) such that the integration of several model-based projects is enabled. 

The project TGG 2.0 (founded by the DFG, in cooperation with the Technical University of Darmstadt) shall yield new concepts and tools for TGGs to integrate modelling processes of several projects. The overall aim is to enable model-driven engineering on a larger scale.

Optimizing models. Search-based software engineering is about solving software development problems by formulating them as optimization problems. A variety of software engineering problems such as software testing, release planning, and software modularization can be considered as optimization problems. As structures in these problem domains are often too complex to be effectively captured with traditional representations in search-based systems, domain-specific models may be the right means of representation. We optimize models by using genetic algorithms.

Zuletzt aktualisiert: 30.07.2019 · struebed

Fb. 12 - Mathematik und Informatik

Fb. 12 - Mathematik und Informatik, Hans-Meerwein-Straße 6, D-35032 Marburg
Tel. +49 6421/28-21514, Fax +49 6421/28-25466, E-Mail: dekanatfb12@mathematik.uni-marburg.de

URL dieser Seite: https://www.uni-marburg.de/fb12/arbeitsgruppen/swt/research

Impressum | Datenschutz