The Expertool Platform is particularly well suited for applications in Translational Medicine, Diagnostic Decision Analysis and Personalized Care, since it is optimized to aggregate disparate content, and to establish rules for interpretation and rationalization. Integration is modeled using coexisting relationships that can be either deterministic or stochastic.
Unlike traditional expert systems that access various content silos, the Expertool architecture's unique homogenous storage and parsing of diverse knowledge types allows heuristics to interact with qualitative and quantitative data. For example, a model including epidemiological, regulatory, clinical trial and life sciences content would not be constrained by the use cases associated with each content source platform (or by internal architecture components), but would enable the discovery of impacts and relationships that had not been previously defined.
Knowledge Science and Engineering
There are many initiatives seeking to address the explosion of data and information with the goals of developing technology to store and manage the volume and defining formal methods for parsing the content and producing findings. Our goal has been to create an environment where the findings can be rationalized and integrated, as well as analyzed as complex systems.
The Expertool modeling architecture is totally abstract, and therefore without disciplinary bias, and supports the definition, coexistence and interaction of an unlimited array of formalisms, taxonomies and ontologies. The Expertool knowledge architecture manages semantic, procedural and episodic content using a homogenous data structure and overlapping methods. As a result, discovery is enabled and analysis is not subject to the scope constraints of traditional modeling tools.
The powerful and precise modeling methods include the following types of links:
Inherent - always relevant within the scope of the model Potential - relevant based on user decision Context-dependent - deduced from the combination of all user and system selections Attribute-driven - dependent on the value of one or more attributes of the selected nodes (including range or substring) and the links change if updates from the source data change the attribute value Formula-driven - dependent on the output of a formula and the links change dynamically as the user navigation of the model changes the relevent values
