In Silico Hypothesis Discovery - Translational Informatics: Realizing the Promise of Knowledge-Driven Healthcare - page 127

Knowledge Base

Defining Actions
Based on
Context

Guidelines or Algorithms
(Procedural Knowledge)

Execution Engine

User Interface

Connecting
Information and
Actions To Create
Actionable
Knowledge

Application
Logic
(strategic Konwledge)

Contextualizing
Patient Data
(Information)

Findings and Diagnoses
(Conceptual Knowledge)

Patient-derived
Data

Fig. 8.4 Overview of a prototypical CDSS platform, incorporating conceptual, procedural, and
strategic knowledge types in order to generate actionable knowledge from patient-derived data

ings, diagnoses and the relationships between them ( conceptual knowledge ); (2) a
set of guidelines or algorithms used to reason upon the preceding database ( proce-
dural knowledge ); and (3) a formal defi nition of the logic used to operationalize the
preceding two knowledge collections ( strategic knowledge ) (Fig. 8.4 ).
It is only when these three types of knowledge are combined that it is possible
to realize a functional decision support system [ 22 ]. Given the close similarities
between such CDSS and the previously introduced framework for in silico hypoth-
esis discovery methods or tools (as is illustrated in Fig. 8.2 ), this phenomenon is
important to keep in mind for the remainder of this chapter.

8.2.1 Knowledge Engineering

The core theories and methods that underlie the ability to systematically and rigor-
ously represent conceptual knowledge inform a set of application-level techniques
known as knowledge engineering (KE). The KE process (Fig. 8.5 ) incorporates four
major steps:

1. Acquisition of knowledge (KA)
2. Representation of that knowledge (KR) in a computable form
3. Implementation or refi nement of intelligent agents (e.g., applications that use
formally represented knowledge to reason upon data sets and generate results of
interest to end-users) or applications
4. Verifi cation and validation of the output of those knowledge-based agents or
applications against one or more reference standards.