Question that I get during Systems Federation and Integration chat session (http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2012_03_01).
NicolaGuarino: @Anatoly: what is the ISO 24... enterprise modelling standard you mentioned?
Specifically it was ISO 24744 and this is not about enterprise architecture/modeling, but your intuition is right.
There are several traditions of enterprise ontology (not all, but related to our issue):
1. Enterprise Architecture.
3rd version of Zachman Enterprise Architecture Framework go as far as to have subtitle “Enterprise Ontology™” (sic! Registered as trade mark! http://www.zachman.com/).
There are dosens (may be hundreds) of such an Enterprise Architecture Frameworks that can be regarded as approaches to Enterprise Ontology (ArchiMate is one of a legion): http://en.wikipedia.org/wiki/Enterprise_architecture_framework
Several of this frameworks have strong mention of ontology, e.g. DoDAF (Department of Defence Architectural Framework, http://dodcio.defense.gov/sites/dodaf20/) have IDEAS foundation ontology as meta-model of it (http://dodcio.defense.gov/sites/dodaf20/Ontology1.html).
There exists many others, like DEMO (http://www.demo.nl/) with main book of prof. Jan Dietz that titled «Enterprise Ontology» and based on language-action perspective (http://www.amazon.com/Enterprise-Ontology-Jan-L-G-Dietz/dp/3642067158/). This is a very interesting approach and has many interesting insights. ArchiMate have mention of DEMO as one of the source of concepts (Value and Meaning types of ArchiMate).
2. Situational Method Engineering.
I mention ISO 24744 that is meta-model for software development process. This is very special thing that formally differ from Enterprise Architecture/Enterprise modeling but closely related to it. The name of a discipline is Situational Method Engineering (SME).
If we understand that people with their tools within enterprise works not ad hoc but according to some methodology (like “eXtreem programming” or other agile methodology, or IBM RUP, Rational Unified Process that is not so agile, or PMI PMBoK that is Project Management Institute Project Management Body of Knowledge or any other Body of Knowledge or methodology) then we need meta-model for that aspect of the enterprise. Most interesting that we have more or less general for many enterprises methods of work and very specific (situational) methods that is used within particular My Enterprise. Situational method engineering tell us that we should think about general method as of repository of method components and then compose method for particular enterprise from this components.
It is appear that “method of work” (and very often “method of development” or “software process” in sense of “process of software creation” is paraphrased as “how to produce something” that is almost exact like systems engineering definition (if you add “according to stakeholders requirements, budget and in time”). Therefore SME is known also as methodology for system life cycle modeling. Enterprise (as enabling system) execute system-of-interest life cycle, and this life cycle is modeled with some meta-model (method ontology).
Main concepts of SME almost the same that you can see in enterprise architectures (process/behaviour, workproduct/passive structure, actor/active structure). As I know not too many people thinking about similarity of Enterprise Architectures (that reflect organization of method enactment) and method descriptions (that reflect common patterns of enterprise architectures).
There are several standards of SME, main of them are:
-- ISO 24744 that is most elaborated from ontology perspective (they even introduce powertypes for UML to show method enactment with clabjects). As I know, authors of this standard communicated with Chris Partridge and aware of BORO). It is best but there are no tools for it (except exotic).
-- OMG SPEM 2.0 (http://www.omg.org/spec/SPEM/2.0/) that is very popular because it has Eclipse-based method composer tool that is positioned as “best process authoring tool” (http://www.ibm.com/software/awdtools/rmc/, they sell it for $400. There exists free version with essential function: Eclipse Process Framework Composer (EPF Composer, http://www.eclipse.org/epf/).
-- Open Process Framework (OPF) that is almost private initiative but you can see online repository with more than 1100 method components for systems engineering process (http://opfro.org).
It is very interested that community of adaptive case management (ACM, better start with http://www.amazon.com/Mastering-Unpredictable-Management-Revolutionize-Knowledge/dp/0929652126) regard method repository as a way to organize knowledge about components of a processes that they will compose on-the-fly (enterprise runtime, not enterprise design time). ACM is interesting in regard for its redefinition of enterprise processes (that often regarded as main part of Business Architecture) on the fly. This is ontologically very interesting: we have emerging case life cycle and need have methods of modeling it. I tried to speak with their community about their requirements to enterprise architecture but they failed to support such a communication (this was in adaptive case management community of LinkedIn).
I am now thinking about aligning enterprise architecture modeling methods, SME life cycle modeling methods and ACM with dynamic and emergent nature of contemporary enterprise. Thus Nicola Guarino is absolutely right when he asks about SME standard as if it is enterprise architecture standard.
TechInvestLab is aimed to work in harmonization all this enterprise modeling mess with enterprise modeling language OrgLan (organizational modeling languge) that is ontology-based according principles that I declared in a yesterday letter to Ontolog Summit list (http://ontolog.cim3.net/forum/ontology-summit/2012-03/msg00034.html)
OrgLan is a language for a TechInvestLab PraxOS project that want to deliver method repository of best organizational practices in form that resulting methods can be directly supported by software. We have Russian-language community http://praxos.livejournal.com/ for PraxOS project that have 90 subscribers now.
We have not too much results with OrgLan and PraxOS yet but our .15926 software platform is ready to support it. We at TechInvestLab are more interesting in enterprise engineering DSL than oil and gas engineering DSL (like P&ID).
3. Terminologies and ontologies in assortment.
One of the most interesting is OMG SBVR (http://www.omg.org/spec/SBVR/1.0/). When people form business rules community (http://www.brcommunity.com/) start to develop standard for Business Rules, they discover that they need ontology of enterprise to formulate rules in concepts of this ontology. Eventually they finish SBVR standard that 90% about business ontology and 10% about business rules. This standard not touch IT-solution architecture (Applications and Technology).
Examples like http://www.aiai.ed.ac.uk/project/enterprise/enterprise/ontology.html -- there are too many of them to mention most important. Most of them is interested only to group that developed these terminologies/ontologies.
It is not very interesting because there are no any kind of enterprise theory behind all this “enterprise ontologies”, this is simply “vocabularies” of common words with vague meaning (e.g. “process”, “function”, etc.).
Best regard,
Anatoly