Patterns for Inter-Organisational Knowledge Sharing

Danny Brash, Janis Stirna

Department of Computer and Systems Sciences
Stockholm University/Royal Institute of Technology
Electrum 230, S-164 40 Kista, Sweden
Phone: +46 8 16 16 18
Fax: +46 8 703 90 25
Email: danny@dsv.su.se, js@dsv.su.se

Abstract

To re-use knowledge that has been proven successful in the past is a goal which is often well worth the effort to investigate in detail. The idea of "patterns" as a method for presenting and using this knowledge is gaining popularity within the software engineering community.

In the current EU finansed research project, ELEKTRA, we are currently expanding the scope of the patterns idea to using it as a way of recording organizational knowledge in the form of "best business practice", in accordance with the objectives of the ELEKTRA project.

Up till now, the academic partners in the project have mainly been concerned with the purely presentation and representational issues. These concern how the organizational knowledge will be structured for storage and retrival purposes. This report will present other issues that we are currently discussing. These are related to the nature of the knowledge, its relevance and its usability. A definition of patterns within the framework of the project is given together with factors that affect their implentation. As is already well understand it is, when developing an information system, not the technical details which are the problem, but mostly the social, organisational and communication aspects which tend to cause failures.

Once the issues of representation, definition and procedure have been agreed upon, we will begin with the population of patterns part of an Electricity Sector Industry tool set, which is currently under construction by a software developer that is a member of the project consortium. The next step is the definition of a procedure for defining what constitutes a pattern. The collection of candidate patterns will entail the definition of an explicit list of criteria with which to measure suitability. Once this has been established, the industrial partners will begin collecting patterns. An actual evaluation of how these are actually maintained and used will need to be undetaken as a more long term research interest and are therefore out the scope of this project. However, we will attempt to take into consideration those factors that which we believe will affect pattern usuage in the long-term.