Two decades of Ripple Down Rules research

Debbie Richards; Debbie Richards

doi:10.1017/S0269888909000241

Abstract: Ripple Down Rules (RDR) were developed in answer to the problem of maintaining medium to large rule-based knowledge systems. Traditional approaches to knowledge-based systems gave little thought to maintenance as it was expected that extensive upfront domain analysis involving a highly trained specialist, the knowledge engineer, and the time-poor domain expert would produce a complete model capturing what was in the expert’s head. The ever-changing, contextual and embrained nature of knowledge were not a part of the philosophy upon which they were based. RDR was a paradigm shift, which made knowledge acquisition and maintenance one and the same thing by incrementally acquiring knowledge as domain experts directly interacted with naturally occurring cases in their domain. Cases played an integral part of the acquisition process by motivating the capture of new knowledge, framing the context in which new knowledge would apply and ensuring that previously correctly classified cases remained so by requiring that the classification of the new case distinguish it from the system’s classification and be justified by features of the new case. RDR has moved beyond its first representation which handled single classification tasks within the domain of pathology to support multiple conclusions across a wide range of domains such as help-desk support, email classification and RoboCup and problem types including configuration, simulation, planning and natural language processing. This paper reviews the history of RDR research over the past two decades with a view to its future.

Other Articles By Authors

Two decades of Ripple Down Rules research

Computing Department

Corresponding author: e-mail: richards@ics.mq.edu.au

Published online: 01 June 2009

Abstract: Abstract: Ripple Down Rules (RDR) were developed in answer to the problem of maintaining medium to large rule-based knowledge systems. Traditional approaches to knowledge-based systems gave little thought to maintenance as it was expected that extensive upfront domain analysis involving a highly trained specialist, the knowledge engineer, and the time-poor domain expert would produce a complete model capturing what was in the expert’s head. The ever-changing, contextual and embrained nature of knowledge were not a part of the philosophy upon which they were based. RDR was a paradigm shift, which made knowledge acquisition and maintenance one and the same thing by incrementally acquiring knowledge as domain experts directly interacted with naturally occurring cases in their domain. Cases played an integral part of the acquisition process by motivating the capture of new knowledge, framing the context in which new knowledge would apply and ensuring that previously correctly classified cases remained so by requiring that the classification of the new case distinguish it from the system’s classification and be justified by features of the new case. RDR has moved beyond its first representation which handled single classification tasks within the domain of pathology to support multiple conclusions across a wide range of domains such as help-desk support, email classification and RoboCup and problem types including configuration, simulation, planning and natural language processing. This paper reviews the history of RDR research over the past two decades with a view to its future.

HTML

Acknowledgments

Many thanks to Paul Compton, the father of RDR, for reviewing an early draft of this paper and for the information, mostly gained from his Website, regarding commercial RDR systems. Also thanks to the anonymous reviewers for their extensive comments and guidance over a couple of revisions. Since the list of acknowledgements relevant to this paper would add even more pages, I extend my thanks to everyone named in the references, most of whom I have had the pleasure of discussing their work with them personally. I hope I have represented your research in a way that is acceptable. Thank you for your work in moving RDR forward. RDR research has been generously funded by the Australian Research Council through multiple grants over the past two decades.

http://ww.pks.com.au/

http://www.cse.unsw.edu.au/~compton/commercial_RDR.html

http://www.kmagent.com/

http://www.ucube.net.au/Products/tabid/119/Default.aspx

http://likewize.com/about.html

http://www.eurobizrules.org/Uploads/Files/Sarraf_2c_20Q_2.pdf, accessed February 20, 2008.

http://www.ivisgroup.com/

http://www.microsoft.com/emea/presscentre/pressreleases/rad2006winnerspr_1522006.mspx

http://www.nabble.com/CG%3A-Sonetto-CG-based-rule-and-search-engine-wins-awards-for-Retail-Affiliate-Management-to3086701.html

http://www.yawl-system.com/

http://www.erudine.com/

http://www.erudine.com/downloads/whitepaper_tacit.pdf

http://web.archive.org/web/20051222121140/www.rippledownsolutions.com/page.aspx?id=27

To reflect how rules and cases are ‘worked-up’, the approach was initially called Interactive Recursive RDR (Vazey & Richards, 2004). That name reflected that it shared the goals of earlier separate work into interactivity and recursion, which were described under early systems. Most recently the technique has become known as 7Cs as it supports Collaborative Configuration and Classification of a stream of incoming problem Cases via a set of ConditionNodes, that is, rulenodes, linked to their Classes and associated Conclusions. Here for simplicity the technique is referred to as CRDR.

The implementation of CRDR includes several conclusion types (e.g. stop, getAttribute(attrName’), showfile(fileName), advise(errorCodeNNN), refer(RuleNodeID)) and attribute types (e.g. Text, URI, combo, imageURI, case(), file(), image(), refer()).

Rights and permissions

References (93)

About this article

Cite this article

Debbie Richards. 2009. Two decades of Ripple Down Rules research. The Knowledge Engineering Review. 24:241 doi: 10.1017/S0269888909000241

Debbie Richards. 2009. Two decades of Ripple Down Rules research. The Knowledge Engineering Review. 24:241 doi: 10.1017/S0269888909000241

{{lists.name}}

Two decades of Ripple Down Rules research

Abstract

Rights and permissions

References

About this article

Cite this article

Article Metrics

Access History

Other Articles By Authors