-: A Semantic Web Based Knowledge Representation and Context-Driven Morphing Framework.
Sajjad M. Hussain · 2009
A knowledge-intensive problem is often not solved by an individual knowledge artifact; rather the solution needs to draw upon multiple, and even heterogeneous, knowledge artifacts. Each knowledge artifact may differ in terms of its modality, origin, and format; and may have different functional/operational roles in different problem-contexts. The synthesis of multiple knowledge artifacts to derive a ‘comprehensive’ knowledge artifact is a non-trivial problem. In this paper, we propose a semantic web based knowledge representation and morphing framework K-MORPH that (a) semantically models the knowledge of various knowledge artifacts found in different modalities as ontologies; (b) semantically annotates the heterogeneous knowledge artifacts based on their respective ontologies; (c) represents the domain-specific constraints and specifications for the morphed knowledge, and treats them as a problem-context; (d) defines morphing constructs, to identify problem-specific knowledge components from the entire knowledge artifacts; (e) reconciles related knowledge components; and (f) generates a verified ‘morphed’ knowledge artifact that contains reconciled problem-specific knowledge from multiple artifacts. We discuss the architecture of a prototype medical knowledge morpher to show the need of knowledge morphing in medical domain.