2018-01-23 Meeting notes
- Dennis Wisnosky
Date
Attendees
Agenda
1) Use Case reminder
2) Where we are on our road map.
3) Open Action Items
4) JIRA Issues Review - https://jira.edmcouncil.org/projects/FCP/issues/FCP-4?filter=allopenissues
5) Todays content discussion.
SMIF OWL-UML
SKOS
RDF/S
6) For next week.
Proceedings:
DN to EK Any more progress on DER. EK, no working on LEI and has a PR really to go soon. All Level 1 data done. Still need level 2 representation. DN will be able to see this work including some changes to BE on GitHub. The Legal Entity piece is there. DN When will there be some small gaps closed? The Q1 plan presented at EDMC last week has what we intend to do short term. DN Need to move basisSwap up a level.
DN JIRA has 2 issues that are the same. Need to delete. Missing many definitions in DER.
PR what is the deliverable for the graph thing. Output format or graphics or what? Don't understand the use of individuals. DN The Graph is not part of FIBO. The deliverable will be an operational ontology version of FIBO. PR Wy is it easier to filter on these individuals rather than pure triples? DA The difference is superficial and mainly speed. DN From an operational point of view we would need to filter out many abstractions that are not meaningful. If FIBO had started as and RDF-S model, that would not have been good, but FIBO would look a lot more like this in the WFBO.
Bobbin Are these decompositions. PR No they are abstractions of the property graph, not the property. PR what is this useful for apart from just the graph? DN This is the basis for a data asset inventory Knowledge Graph. Can create a meta data repository on steroids. Can do concept searches and locate concepts in the graph. And, can link up with data assets. WF is developing the ability to click on a link and that opens up a new graph. E.G. An IRS could be totally dissected. PR Adaptive does this now. DN Then Adaptive is way ahead of other vendors. PR Then the graph output is basically another navigator? DN Need to know the term to find it, normally. But, not with this approach. Able to input vanilla IRS and will find FF IRS, for example.
DA The point of searching for terms in the graph could also be done in OWL. Graph gives easier navigation, not fundamental more information. It is already there in the OWL. Pete says there is no more meaning in the Graph. DA says, yes but it is much more clear. This is not an OWL replacement.
DN This is still a work in process. Current challenge is identifying abstractions and reifications so that this is truly an operational ontology. EK to DN Are u keeping track of wha tis overly burdensome and why. Maybe the OWL itself should be simplified. DN sees this as possibly a FIBO - 2 approach. PR What about SHACL? DN WE need to study this.
DA When we show these Graphs, we suppress pseudo ranges and other properties and that makes the model more clear. Need to work with EK to review what should be changed in FIBO based on this work.
Pete brings up SHACL again. DN Will pursue in due time.
DN the Provisional and Informative have some good work. But many gaps. Missing labels and definitions. What to do? DA Development has Release and Provisional and Informative so that there are no undefined references. DA Let's just byte the bullit and eliminate what does not make sense. DN Let's look at provisional and decide what to cherry pick. There are some really weird things in provisional. There is the concept Instrumentality that makes no sense, for example. There are some notorious abstractions. EK Where these things are there, it is where I have not touched. Where I have been there, I already eliminated the nonsense.
DN to DA Can you get something to me by Friday? DA Yes! Nothing that I have seen are difficult. DA and DN will work this Friday. DW concerned that there is a formal review. FND really has participants and should be in this loop.
DA will point al to the code in GitHub that generates the graph DN is using.