Lecture 9 Health Knowledge Representation


Ontologies, Rules, and Other Forms of Storing Knowledge

In order to be useful in decision support systems, knowledge needs to be stored in forms that can be executed by computer. This lecture discusses several forms of knowledge: ontologies, rules, predictive models, literature and guidelines. It also presents simple ways of representing knowledge base in database and how to construct a very simple decision support system in MS Access.




Decision support systems require knowledge in computable forms. Rules, guidelines, or any other forms of instructions can't be written in plain language (i.e., as websites or word documents). In order for computers to understand knowledge, it needs to be coded.
The first part of the lecture briefly reviews different forms of knowledge that is commonly used in decision support systems used in healthcare.

If the above video does not work, download this file: Lecture Part 1.mp4

 





The second part of the lecture describes how to store ontologies and rules in relational database. While the presented methods are simplified versions of what you may find in real systems (particularly for rules), they show how such encoding is done.

If the above video does not work, download this file: Lecture Part 2.mp4

 







The final presentation shows how to build a very simple decision support system inside MS Access. It requires a simple function written in Visual Basic for Applications (VBA)- the code is provided below in resource section.
The presented simple CDSS executes rules stored in database.

If the above video does not work, download this file: Lecture Part 3.mp4

 

Resources