Chapter 12 - IntelliOntoRec: a knowledge infused semiautomatic approach for ontology formulation in healthcare and medical science

Abstract

Linked Open Data (LOD) on the semantic structure of the World Wide Web plays a vital role in deciding the effectiveness of semantic inferencing for making application-oriented decisions. However, when working with safety-critical systems and critical domains that require scientific acceptance from a community like healthcare, the modeling of linked open data must require expert opinion. The linked open data for medical and healthcare as a domain of choice should preferably be semiautomatic with a standard scheme for automatically modeling domain specific sensitive knowledge and human interaction for verification and a community acceptance for healthcare. In this chapter, a strategic scheme that amalgamates semantic mapping of ontologies using a triadic system of semantic-similarity, Simpson's diversity index, and a novel eXplainable AI (XAI) algorithm, which imbibes the concept of fluid thrust has been proposed. The IntelliOntoRec integrates a diverse source of knowledge from medical journals, RDFs, semantic wikis and follows a three-phase verification of knowledge modeled wherein the domain experts can view the automatically modeled knowledge and contribute. The second phase is the honest review by the community of domain experts and integrates knowledge based on manual modeling and opinions of domain experts and incorporates the grey wolf metaheuristic algorithm. The third phase is the rereview an commit, where the knowledge modeled will be finalized. Further, as the modeling and curation of knowledge are successfully achieved, a rule-based axiom induction using heuristics of information is achieved to successfully organize the linked open data as hierarchies before posting it to the linked open data cloud. The ontology modeling is evaluated based on the scoring of the intelligent crowd of experts, and an overall Modeling F-Measure of 98.557583 has been achieved by the proposed IntelliOntoRec.