HOD2MLC: Hybrid Ontology Design and Development Model With Lifecycle

HOD2MLC: Hybrid Ontology Design and Development Model With Lifecycle

Rishi Kanth Saripalle, Steven A. Demurjian, Michael Blechner, Thomas Agresta
DOI: 10.4018/978-1-5225-3158-6.ch052
(Individual Chapters)
No Current Special Offers


Ontologies have gained increasing usage to augment an application with domain knowledge, particularly in healthcare, where they represent knowledge ranging from: bioinformatics data such as protein, gene, etc. to biomedical informatics such as diseases, diagnosis, symptoms, etc. However, the current ontology development efforts and process are data intensive and construction based, creating ontologies for specific applications/requirements, rather than designing an abstract ontological solution(s) that can be reusable across the domain using a well-defined design process. To address this deficiency, the work presented herein positions ontologies as software engineering artifact that allows them to be placed into the position to share the captured domain conceptualization and its vocabulary involving disparate domain backgrounds, that can then be created, imported, exported and re-used using different frameworks, tools and techniques. Towards this end, the authors propose an agile software process for ontologies referred to as the Hybrid Ontology Design & Development Model with Lifecycle, HOD2MLC. To place HOD2MLC into a proper perspective, they explore, compare, and contrast it to existing ontology design and development alternatives with respect their various phases as related to the authors' work and phases in varied SDP models.
Chapter Preview

1. Introduction

An ontology is a compilation of concepts that can be represented as a class composed of attributes coupled with relationships or associations that define interactions between the classes and augmented with axioms which are rules or constrains on classes, attributes, or relationships. Ontologies are promoted as a part of the Semantic Web to attach knowledge to information thereby aiding users (humans and agents) in various ways. Ontologies are heavily utilized in research such as knowledge engineering and representation, domain modeling (Saripalle, Demurjian, & Behre, 2011), database integration (Konstantinou, Spanos, & Mitrou, 2008), natural language processing (Kaihong, Hogan, & Crowley, 2011), etc. For our purposes, we have focused on clinical informatics that involves medical organizations (e.g., private practices, hospitals, etc.) that represent, store patient data (e.g., demographics, diagnosis, vitals, etc.) electronically using standards such as Continuity of Care Record (CCR)1, Health Language Seven (HL7) Standard Clinical Document Architecture (CDA) (Boone, 2011), etc. and share the data in the form of Electronic Health Record (EHR) or Personal Health Record (PHR). As shown in the top portion of Figure 1, EHRs, PHRs, and other health IT systems are brought together by Health Information Exchange (HIE) (Demurjian, Saripalle, & Behre, 2009) (promoted by the HITECH Act2) into a conceptual setting called the Virtual Chart (VC) which provides a common global view for all potential users. For semantic knowledge, each of these systems employs standard medical ontologies such as: Logical Observation Identifiers Names and Codes (LOINC)3 a standard for identifying medical laboratory observations; International Classification of Disease (ICD)4 to hierarchically organize medical concepts such as diseases, symptoms, injuries, procedure, etc., Unified Medical Language System (UMLS) (Bodenreider, 2004) aggregation of medical ontologies and terminology such as ICD, LOINC, SNOMED, etc.; and have a semantic network and metathesaurus. The main objective of our work is to achieve interoperability among the health systems by unifying not only the data, but most importantly, unify their ontologies through their schema or model integration, as shown in the third horizontal box from the top of Figure 1.

Complete Chapter List

Search this Book: