Technology
Business rules
The Sonetto Rules Engine uses a novel algorithm that improves upon the performance characteristics of classic rule evaluation algorithms and has the advantage of not being affected by the number of rules in the system.
The decision-making model, i.e. the rule base, is captured through the ‘Ripple Down Rules’ (RDR) knowledge acquisition methodology. Rules are only added, never replaced, updated or deleted. They are either new, or specialise a previously added rule, empowering business users to easily model their process.
"Ripple down rules"
Ripple Down Rules is a technology for building and maintaining knowledge based systems. The central idea behind Ripple Down Rules is this:
When a domain expert is asked "Why does a certain conclusion apply to this situation?" the expert never explains how the conclusion was reached. Rather, he or she justifies that the conclusion is correct. This justification varies with the context in which it is given.
In most other Knowledge Acquisition approaches, this phenomenon is seen as a practical measure, but incidental to the main Knowledge Base System strategy. In contrast, Ripple Down Rules recognizes it as the central and necessary aspect of any interaction with an expert.
When an expert provides and justifies a conclusion, he or she does this in a particular and specific context. However, there will always be contexts that the expert did not foresee, nor could foresee, that will eventually require this conclusion to be modified, and a new justification given in this new context.
That is, the normal process of Knowledge Acquisition involves continual refinement and correction, not the creation of a set of "universal truths" that apply in all situations. In fact, most knowledge is acquired in what has been traditionally known as the "maintenance" phase.
An explanation of a conclusion, on the other hand, implies that the conclusion will be true in any context. This is a mistaken expectation to have of an expert, and a misunderstanding of the Knowledge Acquisition process.
Knolwedge acquisition
The key to Knowledge Acquisition with Ripple Down Rules technology is as follows:
The process of Knowledge Acquisition begins when the expert identifies a particular case for which the Knowledge Base System has either made an incorrect recommendation, or has not been able to make any recommendation as it may know nothing about this type of case.
This particular case is then stored by the system as a Cornerstone Case, as it is case providing the context for the new conclusion.
The expert is then asked to identify the features in the Cornerstone Case which justify his or her new conclusion about this case.
The system requires the expert to select sufficient features in their justification so that past Cornerstone Cases, for which rules were previously added, are not reasoned about differently by the Knowledge Base System.
Contact Us to learn more about our solutions