Jess 7 Features
Related Web Sites
About This Site
, the Rule Engine for the JavaTM
|Hybrid Expert Systems
Hybrid expert systems employ two or more representations of expertise to emulate reasoning in some domain. Typically, the component expert systems are highly complementary. That is, the weaknesses of one are compensated for by the other and vice versa.
Rule-based expert systems (RBES) -- which you create with Jess -- are very good at making decisions where the domain knowledge is very well structured and defined. Because of this, RBES can be used to reason from first principles to arrive at a general conclusion. RBES tend not to perform well when the domain knowledge is ill-conditioned, poorly-structured, or sparse.
In these instances, RBES can be paired with another expert systems metaphor to overcome that limitation. One particular metaphor that works very well with RBES is Case-Based Reasoning (CBR). A CBR system emulates the very human-like practice of re-using past experiences and problem solutions to synthesize new solutions to new problems. Collections of past problems and their associated solutions are stored as cases. When given some search criteria, a CBR system tries to find the case or cases that most closely match the criteria according to some evaluation algorithm. The most popular algorithm is called Nearest Neighbor. The strength of CBR is that if the adapted solution works, it is rolled back into the case base as new knowledge -- therefore the more a CBR system is used, the more it learns.
RBES + CBR Example
As an example, in a hybrid expert system employing both RBES and CBR components for mechanical engineering design, a user might request assistance in determining the failure modes of a new machine part. The system would first try to locate a similar part in the case base, which then might be automatically adapted by the system to form a new solution, or manually adapted by the user. If no usably similar part can be found, then the system would switch to the RBES to reason from its knowledge of failure modes and the given part parameters to predict the part's failure modes and critical loads.
In this way, the entire system functions more like a human expert than if powered by a RBES or CBR component alone. Obviously, hybrid expert systems are very complex to build. However, with the increasing need to manage corporate knowledge, such systems can readily find applications in e-commerce, design, decision support, and e-learning.
RBES can be classified as either forward-chaining or backward-chaining, depending on how the rule engine makes its inferences. If we relax the pedagogical notions of hybrid expert systems put forth so far, we can have both forward and backward chaining aspects of RBES in the same application. For example, let's say that we had a design system again where the initial inputs were sparse or information was incomplete. We could have a backward-chaining sub-system ask questions to the user to elicit design parameter information that would channel the design process toward a more defined and limited subset of all available alternatives. Once those parameters had been established, the forward chaining sub-system would take over and direct the user to a final concluding design. A more concrete example would be a system for choosing object-oriented design patterns that queried the user about her intent and needs using backward-chaining, then used that information to reason which design pattern was most applicable using forward-chaining.
Feel free to add anything!
Front Page |
Recent Changes |
Powered by Friki |
Last Edited: 16 November 2005