Acquiring Expert Knowledge for the Design of Conceptual Information Systems

Abstract

Conceptual Information Systems unfold the conceptual structure of data stored in relational databases. In the design phase of the system, conceptual hierarchies have to be created which describe di erent aspects of the data. In this paper, we describe two principal ways of designing such conceptual hierarchies, data driven design and theory driven design, and discuss advantages and drawbacks. The central part of the paper shows how Attribute Exploration, a knowledge acquisition tool developed by B. Ganter can be applied for narrowing the gap between both approaches.

Service integration in distributed control systems: an approach based on fusion of mereologies

“…Let us consider the formal context C =( Φ ,G,F), where G denotes the set of relevant particular sub-goals and F ⊆ Φ×G, a binary relation which holds between physical entities and particular sub-goals. As suggested in [20], the set of subgoals is extended with hierarchical conceptual scales such as the intent includes sub-goals, goals (i.e., services) and the instrument scale (highest level). Higher level scales define a partially ordered set (G, ) provided that the set G contains exactly the minimal elements of G. Hierarchical conceptual scales are filled according to information input by the user concerning goals definitions.…”

“…The goal mereology is derived from the subsumption hierarchy of conceptual scales where the many-level architecture of conceptual scales [20] is extended taking into consideration the mereological nature of the extents. Higher level scales which relates scales on a higher level of abstraction provide information about hierarchy and help to derive a hierarchy like the mereology.…”