John Kingston AIAI, Edinburgh University Knowledge management is “the identification and analysis of available and required knowledge assets and knowledge asset related processes, and the subsequent planning and control of actions to develop both the assets and the processes so as to fulfill organizational objectives.” [2] The above definition of  knowledge management implies that is necessary for organizations: to be able to capture and represent their knowledge assets;  to share and re-use their knowledge for differing applications and differing users; this implies making knowledge available where it is needed within the organization; to create a culture that encourages knowledge sharing and re-use. From an information technology viewpoint, the true challenge for knowledge management therefore lies in accurate capture and accessible representation of knowledge assets. These are the areas to which the science of knowledge engineering has the most to contribute. For a knowledge asset to be represented in a manner which is accurate, complete, embedded in its context, and yet comprehensible, multi-perspective modeling is required. As the name implies, multi-perspective modelling requires that a knowledge asset should be represented using a collection of knowledge models, each of which takes a different viewpoint on that knowledge. The diagram formats may (and probably should) differ between perspectives, but all knowledge items are drawn from a single underlying repository. We propose that the various perspectives that are recommended by the CommonKADS methodology can be summarised under the following headings: how a process is carried out, who does it, what information is needed, where that information comes from, when each activity must be carried out, and (less explicitly) why the process is performed. Table 1 gives more detail on the expected contents of these perspectives.   Perspective Description What Declarative knowledge about things as opposed to procedural knowledge about actions. “What” knowledge encompasses concepts, physical objects, and states. It also includes knowledge about classifications or categorisations of those states. How Knowledge about actions or events. It includes knowledge about which actions are required if certain events occur; which actions will achieve certain states; and the required or preferred ordering of actions. When When actions or events happen, or should happen; it is knowledge about the controls needed on timing and ordering of events. Who The agents (human or automated) who carry out each action, and their capabilities and authority to carry out particular actions. Where Where knowledge is needed and where its comes from -- communication and input/output knowledge. Why Rationale; reasons, arguments, empirical studies and justifications for things that are done and the way they are done.  Table 1: Descriptions of perspectives © 2000 SGES Last updated January 3rd 2000