Production Rule Representation

The Production Rule Representation (PRR) is a proposed standard of the Object Management Group (OMG) to provide a vendor-neutral rule-model representation in UML for production rules as used in forward-chaining rule engines.

History

The OMG set up a Business Rules Working Group in 2002 as the first standards body to recognize the importance of the "Business Rules Approach". It issued 2 main RFPs in 2003 – a standard for modeling production rules (PRR), and a standard for modeling business rules as business documentation (BSBR, now SBVR).

PRR was mostly defined by and for vendors of Business Rule Engines (BREs) (sometimes termed Business Rules Engine(s), like in Wikipedia). Contributors have included all the major BRE vendors, members of RuleML, and leading UML vendors.

Evolution
  1. The PRR RFP originally suggested that PRR use a combination of UML OCL and Action Semantics for rule conditions and actions. However, expecting modellers to learn 2 relatively obscure UML languages in order to define a production rule proved unpalatable. Therefore PRR OCL was defined that included OCL extensions for simple rule actions (as well as external functions). PRR OCL is currently considered "non-normative" i.e. is not part of the PRR standard per se. PRR beta applies just to a PRR Core that excludes an explicit expression language.
  2. The PRR RFP envisaged covering both forward and backward chaining rule engines. However, the lack of vendor support for / interest in backward chaining caused this to be revise to forward chaining and "sequential" semantics. The latter is simply the scripting mode provided by many BPM tools, where rules are listed and executed sequentially as if programmed. This provides PRR with better compatibility with typical BPM scripting engines (and acknowledges the fact that most BREs today support a "sequential" mode of operation, improving performance in some circumstances).

Status

PRR is currently at version 1.0. Per the OMG process it is currently being revised by a Revision Task Force expecting to release a 1.1 version in 2010.

Revision task force members were ILOG Inc (co-chair), NoMagic Inc, TIBCO Software Inc, Business Semantics Ltd, Inferware LLC, Sandpiper Software Inc, and 88 Solutions Inc.

gollark: It would also not be very useful for spying on people, since they would just stop saying things if they got a notification saying "interception agent has been added to the chat" and it wouldn't work retroactively.
gollark: One proposal for backdooring encrypted messaging stuff was to have a way to remotely add extra participants invisibly to an E2Ed conversation. If you have that but without the "invisible" bit, that would work as "encryption with a backdoor, but then make it very obvious that the backdoor has been used" somewhat.
gollark: Not encryption itself, probably.
gollark: They don't seem to want to *ban* end-to-end encryption as much as backdoor the popularly used stuff. Which is still bad. I should finish writing that blog post on it some time this decade.
gollark: It's probably with consent to the extent that *any* social media apps do, i.e. "the long incomprehensible privacy policy says we can".

See also

Sources
  1. OMG Press Release on PRR Adoption Dec 2007
  2. OMG Specification page
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.