An Automatic Knowledge Acquisition Mechanism for Independent Inference Engine Module of CDSS

Clinical decision support system (CDSS) has been increasingly applied to help clinicians analyze patient data and make clinical decisions at present. The kernel part of CDSS is inference engine (IE) module whose independence is very important, which separates the knowledge reasoning from knowledge representation and causes them to turn two independent processes while most of IE modules currently used haven´t achieved this goal yet. IE module consists of two parts: an independent IE and the automatic knowledge acquisition mechanism for IE. In this paper, CLIPS is chosen as independent IE and we devoted to establishing the automatic knowledge acquisition mechanism using two methods proposed by us which were named Multi-Rule Sets Abstracting and Single-Rule Series Abstracting. Via comparison, the latter was demonstrated to be more effective for purpose. In practice, an experiment on IE module indicated that this automatic knowledge acquisition mechanism realized its independence.