Chapter 7 - Probabilistic, syntactic, and semantic reasoning using MEBN, OWL, and PCFG in healthcare

Abstract

With the advent of state of the art of artificial intelligence technologies like deep learning, several digital healthcare records are yet to be fully explored. Several data processing and analytics methods treat the data as numbers and strings and attempt to make a sense out of it using data mining and machine learning techniques. The current data processing and data analytics methods are not new to handling electronic healthcare records, yet they suffer from a typical problem of missing out the semantic information associated with the healthcare records which humans excel at. Semantic knowledge bases in the form of ontologies and the insightful syntactic patterns in healthcare data, especially clinical text documents, could be of great help for several medical research projects aiming at developing intelligent semantic reasoning methods for supporting healthcare decision support systems. The chapter thoroughly introduces the key concepts and terminologies in syntactic and semantic reasoning using Multi Entity Bayesian Networks (MEBN), Web Ontology Language (OWL), and probabilistic context-free grammars (PCFG) with a special focus on applications to the healthcare domain.