Research

Over the past decade, architectural patterns have become an integral part of software design practices. Architectural patterns provide guidelines for modeling the structural and behavioral aspects of a system design. Architectural patterns also provide means to reason for the quality attributes of a software system and help to document the design decisions taken by an architect. Regrettably, describing, finding, and applying architectural patterns effectively in a system design remains largely ad-hoc and idiosyncratic. Our work in this domain focuses on both the theoretical and practical research directions. On the theoretical part, we have published several papers in renowned journals and conference, while on the practical side we are working on providing tool support for effectively modeling patterns in a system design. Our ongoing research work is focused on analyzing the impact of architectural patterns on quality attributes, devising an approach for effectively modeling pattern variability, documenting design decisions, mining patterns in a system design etc.

In the context of a large industrial project with Philips Healtchare, we aim at developing methods and tools for optimizing system evolvability, i.e. the ability of a system to evolve easily in the face of changing requirements. Our research is performed in an empirically manner exploring forward and reverse engineering processes to construct and maintain architecture views. In particular, we have developed a method to collect dynamic information and build execution architecture views of an MRI system, a real large and complex software-intensive system. The architecture views constructed or recovered using our method, provide architectural information about the actual MRI system behavior. This information is useful for architects and designers to understand the actual MRI system behavior, conduct dependency analysis activities, and identify the impact and propagation of changes, which is important to efficiently response to changing requirements.

There is a growing awareness of the importance of Architectural Knowledge (AK) in the software architecture community. Within the context of the Griffin project, we investigate this notion of AK and do this in the context of an envisioned architectural knowledge grid. This knowledge grid is an omnipresent knowledge infrastructure for capturing, publishing, sharing, and managing architectural knowledge. So far, we have investigated what AK entails, how this can be captured in architectural documentation, the relationship between architectural analysis and AK, and architectural decisions, which are an important form of AK. Currently, we are investigating the AK sharing and reusing across organizations, especially in geographically distributed context using various AK domain models. The objective of this research is to (1) predict the AK sharing quality more accurately, (2) facilitate the AK sharing activity before real sharing cost was spent, and (3) take the user involvement into account for the AK sharing in different sharing context, and (4) carry out and evaluate AK sharing AK in industrial organizations.

Software As Service for the varying needs of Local e-Governments is a research collaboration between the University of Groningen, Cordys (technology partner), and the Government Projects EGEM and WWZ (Cooperation between 24 municipalities in Groningen on the WMO law). The project aims to find solutions to an acute problem of the Dutch municipalities over the past years: effective e-Government. The municipalities need to provide more and more digital services to the citizens. As a result they need to develop numerous and complex software systems to implement these e-services. The SAS-LEG project proposes to use the Software as Service (SaS) principle in order to implement the law once and offer it as a customizable service to several municipalities. The project will conduct research aiming to produce a software engineering approach to architecture such a SaS. The expected benefit for the municipalities is reduced implementation and evolution costs in terms of both time and money. The technology partner also aims at acquiring the know-how and technology to lead to commercial products for SaS-based e-Government at the end of the project.

More information about the completed research projects by SEARCH group.