Computer Aided/Assisted Software/Systems Engineering (CASE) have become largely a de facto standard for many system development initiatives due to advances in computing power/cost, object-orientation an more flexible services, client-server architectures and graphical interfaces. Another reason is that system development is still plagued by mediocre productivity and insufficient quality, which are expected to overcome by task, process, and coordination technologies (CASE). These issues raise several research questions that have remained on the agenda of system development researchers over the last decade.

The first question deals with the need to describe, analyze and manage multiple types of descriptions (design artifacts) in the CASE environment in a way that allows for their consistent, efficient and user-friendly manipulation. This has raised the issue concerning how to specify these descriptions and their manipulation requirements. Therefore we have been in the past decade studied extensively what are the functional and cognitive requirements for representation schemes to adequately describe and model the metadata stored, represented and manipulated in the IS repository.

The second question deals with the need for using a disciplined process in using and applying these representation schemes. This question we call the method engineering problem: how to specify, manage, develop and maintain a repository schema and the functionality to specify and manage software artifacts? Both these issues have become much more widely researched during the recent years, and both commercial CASE tool builders and methodology vendors are suggesting more disciplined solutions to this problem (like UML specification in UML and associated UML generic process model).

The third question concerns the need to develop computer tools for method engineering. This need is necessitated by the need to develop methods fast and to improve the quality of method use. Both of these demand heavy computerization. We call this area computer aided method engineering (CAME).

The fourth question deals with the future development methods. While current CASE tools support methods that were developed either in the 1970's (structured methods), or nowadays mostly in the early 1990's (object-oriented methods) they are mainly based on manual methods. Accordingly their support functionality does not take into account the versatile functionality a computer-supported environment can offer, for example hypertext or visualization features. Therefore the method modeling issues must also be explored from the viewpoint of how to develop "computer engineered" methods instead of remaining caught in a "paper-and-pencil" mindset.

MetaPHOR Home Page
Pages created by Steven Kelly. Last update September 24, 1999 by Jouni Huotari.