Here’s a quick addition … The RDF representation uses the names of the classes and predicates in the URIs that identify them. This means that when the names change, the URIs change and this invalidates all of the previous uses. As the SIG considers only the number to be important, there is a mismatch of expectations around persistence and versioning.
Examples: E78_Collection versus E78_Curated_Holding and the recent thread about renaming translation_of. Rob From: Crm-sig <crm-sig-boun...@ics.forth.gr> on behalf of Richard Light <rich...@light.demon.co.uk> Date: Wednesday, January 17, 2018 at 3:46 AM To: Jim Salmons <jim.salm...@factminers.org>, "crm-sig@ics.forth.gr" <crm-sig@ics.forth.gr> Subject: Re: [Crm-sig] ISSUE Recording an E41 in RDF Jim, Thank you for the encouragement. I have put the document in its current form at: https://docs.google.com/document/d/1zCGZ4iBzekcEYo4Dy0hI8CrZ7dTkMD2rJaxavtEOET0/edit?usp=sharing and it is editable by anyone with the link. As you'll see, there is little that is new in there (although there might already be things to argue about!), but there is the outline of a more substantive document. All suggestions and contributions gratefully received. Richard On 16/01/2018 23:42, Jim Salmons wrote: Richard and SIG members, On 16/01/2018, Richard Light wrote [rest of thread snipped for brevity]: “I have started an "issues with RDF" document, but on reflection it may be more constructive to make it into a first attempt at the guidance I am asking for. I'll spend this afternoon pulling together material which I can easily find (e.g. the introductory comments in the RDF Schema document), and see what questions that exercise answers.” The recent flurry of conversation relating to the interplay of #cidocCRM and #RDF is most interesting and timely, both to me personally and, I believe, to the larger SIG mission of championing our model’s utility to those who are interested but hesitant to explore and adopt it in practice. == On the "Big Picture" Community Level... == 1. Richard, I would be very interested to see your working document mentioned above as soon as it is available and would love to be involved in its draft evolution as I would qualify as a highly-motivated non-expert reader with good writing/editing skills. 2. I know that this mailing list is very focused on the "tight" conversations of core and significant modeling issues and their resolution. Given that wrestling with "#cidocCRM in #RDF" is itself a gnarly domain that will likely engender its own level of detailed conversation, and given that the SIG is currently having an in-person meeting on current issues and future directions, might it be appropriate, via the energy and interest at the current meeting, to form a Working Group on this topic and spawn its own mailing list with a charter to explore this topic and come back to the full SIG with draft documents (e.g. the afore-mentioned "primer") and recommendations in response to its charter? If such a working group were to be formed, I would very much like to be involved. Putting on my "marketing hat" for a moment, I believe that the better we address #cidocCRM in #RDF, especially in terms of practical and example-based documentation and learning materials, that this will be the most important initiative we can take at this time to advance the adoption of the #cidocCRM in deployed and new #LOD systems/collections. Happy-Healthy Vibes to All and a Happy New Year, -: Jim:- www.researchgate.net/profile/Jim_Salmons<http://www.researchgate.net/profile/Jim_Salmons> www.medium.com/@Jim_Salmons/<http://www.medium.com/@Jim_Salmons/> (my #CognitiveComputing/#DigitalHumanities articles) P.S. As a postscript, I provide these comments with regard to my own personal learning and research experience... == Optional on my Personal Interest in #cidocCRM & #RDF == At a personal level, some in the SIG know that I am a U.S.-based independent (and untrained) #CitizenScientist working my post-cancer #PayItForward Bonus Rounds to contribute my best efforts at the intersection of #DigitalHumanities and #CognitiveComputing. As a “software guy” I spent the bulk of my career as a Smalltalk developer and was particularly active during the initial wave of the software patterns movement. I was drawn to the #cidocCRM through my desire to apply ideas for metamodel-driven design of “self-descriptive executable model” frameworks from my prior Smalltalk work. I want to apply these ideas to my research that takes advantage of the emerging technology of graph databases. As a “pure OOP” Smalltalker, I had a “knee-jerk” reaction of disinterest in #RDF as its level of detail in notation reminded me too much of what we “pure OOPers” felt about the object-orientedness of C++ and Java. I have been using Neo4j’s property graph database for my initial applied research but lately became disenchanted with it. As I surveyed my technology-provider options, I decided that my piqued interest in Linked Open Data warranted a reevaluation of #RDF and the available triple store products as a means to pursue my work in development of the MAGAZINE #GTS (ground-truth storage) format based on a #cidocCRM/FRBRoo/PRESSoo ontological “stack.” I am now fully committed to redirecting my #cidocCRM-based research platform around #RDF (along w/ #TEI) primarily for these three reasons: * I found Ontotext's GraphDB to be an excellent company and technology, both in its principal product and in its all-important documentation, self-driven learning resources, and its helpful tech support community. * Once I was "bitten" by GraphDB, I began an intensive effort to come up to speed on #RDF through self-study and found the most incredibly-written and super-helpful book, "Semantic Web for the Working Ontologist: Effective Modeling in RDFS and OWL, 2nd Edition" by Dean Allemang and James Hendler (book companion website http://www.workingontologist.org). * My interest in software patterns led me to Pascal Hitzler (http://www.pascal-hitzler.de/) and the ODPA, the Association for Ontology Design & Patterns and their website at http://ontologydesignpatterns.org with associated Google group mailing list at this shortened URL https://goo.gl/x6MJjM. Through my initial involvement in this community, I am excited to note that I will be attending #us2ts, the 1st U.S. Semantic Technologies Symposium in early March in Dayton, Ohio. Of course I will be bringing my interest in ontology design patterns and the #cidocCRM to this event which is geared toward developing a North American cross-discipline semantic technologies research community. More information on this event is here http://us2ts.org/. Finally, I am also pleased to note that as part of my #PayItForward Bonus Rounds I served on the Program Committee of #DATeCH2017 and my fellow cancer-survivor wife and I had two papers accepted for a poster at this event, a PDF of which is available here https://1drv.ms/b/s!AtML1v0eUlpEgoAJ_FH6CMU5luOUBA. To those who read this optional postscript... another Happy-Healthy Vibes, -: Jim :- . -- Richard Light