This is a white/technical paper on a cutting edge expert system, the Health Information Technology Development Resource (HITdr). It outlines the concepts involved with problem-solution pairings, and employs an inference mechanism that uses similarity-based matching algorithms to select recommended solutions given user input. The project develops a knowledge-based system that serves as an online development resource of “lessons learned” for health information technology system developers. Further, an ontology is defined to support the consistent representation of knowledge within the system as well as to enhance information retrieval.
A system is proposed that will provide access to the invaluable knowledge gained through actual EHR implementation experiences. Given the unstructured nature of the problem coupled with the existence of prior knowledge scenarios and domain-specific concepts, a case-based reasoning (CBR) system supported by a semantic framework (i.e., an ontology) is recommended. Not knowing what might work best in a given context – the unknown – makes it difficult for system developers to propose and justify new technologies. Cost is another factor. Finding a way to avoid “reinventing the wheel” with each new implementation is an important step towards reducing both uncertainty and cost.