Knowledge representation plays a vital role in reasoning which is a part of the Decision Support Systems (DSS). Medical ontology can be a model of the information from a clinical domain that will contain all the relevant ideas related to the patient information, treatment, their medicine and clinical procedures. Ontology allows knowledge inference and reasoning. The benefits of applying ontology for information illustration are: information sharing, standardization of medical terms and support for automatic semantic reasoning (Wang et al, 2004). In this manner, ontologies integrate descriptive, actionable and factual information and they can be populated using the PIKES approach. (Francesco et al, 2016)
In this chapter, our aim is to discuss different knowledge representation frameworks, manipulation of data and metadata and their storage (Petar, 2016). Representation and Inference must be the core of any knowledge representation systems. So here, we have two Ontology representation mechanisms – namely RDF and OWL Frameworks. Rules and query languages for handling such metadata repositories are discussed in the below sections.
RDF is a W3C standard (RDF Working group, 2004) for describing metadata and modeling the resources in the Web. It is written in XML format. Any kind of knowledge available in RDF can be decomposed into triples. Semantics and rules can be applied to relate the knowledge based on their meanings (Chakkarwar et al, 2016) and with their timestamps (Barbieri et al, 2010). RDF triples are a combination of three variables- Subject, Predicate and the Object. The objects are always literals and the subjects and predicates are resources. Resources are identified by Uniform Resource Identifiers (URI) when the data is from a Web repository and a set of statements when the data is from text or relational databases. For instance, <http://en.wikipedia.org/wiki/Neurotransmitter#> is a triple communicating that the subject Dopamine functions as a Neurotransmitter which is an object. URL (Uniform Resource Locator) has been utilized to find a Web page we need to get to. RDF uses two unique sorts of URI to recognize a given resource, in particular slash URI and hash URI. URI using slash is basically an ordinary URI that we all know and URI using hash is used as below:
Ordinary URI + # + fragment identifier
URI using hash and slash to recognize a Sick person on the Web can be used as mentioned below.
http://www.clinic.org/Sickpersons/Sickperson1
http://www.clinic.org/Sickpersons#Sickperson1
The ordinary URI for hash URI is given by http://www.clinic.org/Sickpersons/ Sickperson1and its fragment identity is given by # Sickperson1. An RDF file must have the below format.
The RDF root element is <rdf:RDF> which defines that an RDF document is referenced with a namespace xmlns. <rdf:Description> gives description about the resources to be stored in the RDF file.
<rdf:RDF
xmlns:rdf=”http://www.w3.org/1999/02/22-rdf-syntax-ns#”>
<rdf:Description
...Description of resources
</rdf:Description>
</rdf:RDF>
Example 1: A simple set of statements describing the hill station Kodaikanal is considered. The area of Kodaikanal is 21.45 sq-km. Its altitude is 2133 metres. The food crops in Kodaikanal are plums, peaches and pears. The locations (place of interest) of Kodaikanal are stored in URI http://www.hill-stations-india.com/kodaikanal/. Converting this statement into triples (Subject –Object –Predicate) will take the form as in Table 1.
Table 1. The below statements gives the RDF representation for the above Example.
<rdf:RDF
xmlns:rdf='http://www.w3.org/1999/02/22-rdf-syntax-ns#'
xmlns:NS=”http://tamilnadutourism.org/HillStation/HillStationMain.aspx” >
<rdf:Description rdf:about ='Kodaikanal'>
<NS:Area> Area 21.45-sq-km </NS:Area>
<NS:Altitude> Altitude 2,133 metres </NS:Altitude>
<NS:Foodcrop> peaches, plums, pears </NS:Foodcrop>
<NS:Location rdf:resource=”http://tamilnadutourism.org/HillStation/TNHillStations/HS_Kodaikkanal.aspx?catid=030101P01 “/>
</rdf:Description>
</rdf:RDF>
Representing the information in the Web using RDF in the above format will help the user group the information based on the subjects and predicates. BM25F ranking function for RDF data, can also be used in RDF searching and organizing data.(Blanco et al, 2011) To ensure that the RDF statements are valid, they can be validated through the W3C RDF validator. The results are shown in Figure 1.