Managing Patient Record Instances Using DL-Enabled Formal Concept Analysis

Abstract In this paper we describe a general logic-enabled Formal Concept Analysis (FCA) approach to manage patient record instances. In particular, the conceptual model of the domain is represented as a breast cancer imaging ontology using a Description Logic (DL)-based web ontology modelling language, DAML+OIL. Patient records are treated as instances with regard to the ontology. We studied a knowledge base (KB) with 1,500 anonymous cases (2,200 abnormality instances), whose routine management functionalities, e.g. instance retrieval, instance introduction, KB visualisation and navigation, are driven by a DL-enabled FCA engine. We demonstrate that our approach is capable of conveying not only the syntactic but also the semantic information, presenting direct visual correlations between logic formulas (intent) and instants (extent) in the knowledge base and facilitating a user-friendly graphic interface easing the knowledge management processes for people with limited expertise on kno...

• Type: Conference
• Year: 2004
• Where: EKAW
• Authors: Bo Hu, Srinandan Dasmahapatra, David Dupplaw, Paul H. Lewis, Nigel Shadbolt
• Keywords: EKAW 2004, Information Retrieval, Knowledge Base, Patient Record Instances, Patient Records