The DeSyGNER knowledge management architecture: a building block approach based on an extensible kernel

  • Authors:
  • Robert A. Greenes;Stephan R. A. Deibel

  • Affiliations:
  • Decision Systems Group, Harvard Medical School, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA;Decision Systems Group, Harvard Medical School, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA

  • Venue:
  • Artificial Intelligence in Medicine
  • Year:
  • 1991

Quantified Score

Hi-index 0.00

Visualization

Abstract

The Decision Systems Group has been developing a 'building block' approach for creating Knowledge Management (KM) applications for medical education and decision support. Potential functions and knowledge access modes to be supported include query, browsing, testing, simulation, didactic instruction, problem solving, and personal file management. Knowledge is considered to be available in multiple forms, non-adaptive and adaptive. We believe that organization and combination of disparate components, in order to build varied and complex applications as required for KM, is best achieved through a software engineering approach based on a kernel set of functions that provide a consistent set of services for all applications, facilitating extensibility and inter-application compatibility. For this purpose, we are exploring a prototype kernel architecture called DeSyGNER (the Decision Systems Group Nucleus of Extensible Resources). Features addressed by DeSyGNER include methods for decomposition of applications into modular units and identification of their functional dependencies; methods of structuring applications to separate their storage, processing, and presentation components; database requirements for indexing and composing complex structures from disparate, disjoint data elements; and methods to support multi-user cooperative development.