allValuesFrom means the thing may only be one or more of those but could be none;
someValuesfrom means that all uses of that property for that restriction must include at least one f those things, but may also have others.
Cardinality is more specific as to numbers and the meanings if these are clear
Discussion items 20150804 FIBO-BE FCT.docx
20150804 FIBO-BE FCT
BE59 proposals to partition certain things more crisply. Relates to the kinds of trusts, functional entities in general. Relates to the reconciliation on the legal forms as well.
BE58 DN view to get of attorney on trusts and which are functional entities in progress.
BE56 MB to raise JIRA issue this is the JIRA issue. In progress.
BE55 similar. This is a simple name change.The action is to write up the proposed solutions in the JIRA issue machinery now that we have this completed.
BE54 is the last piece for legal forms we have to reconcile. DN still finding complexity in governments and governmental agencies.
BE53 covered Aim is to show connection between LEI and jurisdictions.
BE52 rationalization of Business Entity. This is work we need to do for OMG.
BE51 natural person. Likely we will be able to close this out. Natural Person has been reviewed and has a new name, Legally Capable Person, which is in Foundations.
BE50 names of ownership and control properties. We did not get to that yet, this will be the next major sub domain of BE we needed to get our arms around now that we have substantively completed n legal forms.
May need another such meeting after the legal forms changes are completed. FBC may have dependencies on some of these constructs, and they would also need to
know of name changes. So we would need another such meeting to deal with the impacts on FBC. That would be after this work on legal forms is promoted to pink.
Elisa: Not concerned about that, since her work depends on the last published version of the BE spec. This is the version frozen in March, and the FBC spec has to be written describing any changes or relations to the March version. This is because we are publishing FBC in 2 week from today. So any changes that would subsequently impact FBC, will require us to later open issues in FBC to do the reconciliation work David is describing. This is not in the critical path for the Aug 24 submission. OMG process does not require this since until the BE FTF report is presented at OMG (presumable December!), the FBC spec has no dependency on that upcoming version.
DN Will there be an unintended consequence on BE from the FBC work? David suggests w hold FBC back so we can move them forward together.
EK: The OMG process is not that brittle, and does allow multiple versions of OMG specs to exist in parallel. Until the FBC RFC is approved and has gone through the FTF process itself, there is less of an issue (we can make changes to FBC as we go). Also the process allows multiple versions anyway. What we will specify in FBC is the dependency on a specific version of BE as of a certain date. Then DN and the BE FCT makes changes to the BE spec. Then if people want to use FBC they would remain dependence on the version that was given as the version it was dependent on in the written spec itself. Then there would be two version of BE that people can look at, the March version and the later version that the BE FCT promotes. So therefore these dependencies are not really an issue, an we don't need to hold back any version of FIBO to take account of changes in another version of FIBO.
DN: “Glad to hear this.”
So, as we promote BE forward, later revisions of FBC and others would be revised to
point to the new content in the later BE as they require. FBC will depend on the prior version of BE until and unless they opt to update it.
EK: Meanwhile BE has so many changes that it is no longer the same specification and the AB may push back on this on the basis that the changes are quite considerable. The AB may encourage the submitter to put forward a version 2. Unless we are able to show a documented logical progression from what was published in March to what is being submitted now if the AB are comfortable with that we may be OK.
MB notes that the changes from March are now less extensive, following the work he
and David did recently.
BE45 Done, see above. have renamed ("eliminated") natural person
BE44 partial lattice patterns. This needs to be coordinated with the work by Foundations FCT on their 2 open issues that relate to ownership and control. The two FCTs will work together on these.
BE43 Benefit Corporation. Will look at types of corporations.
BE42 improper model of registered Address.
BE40 Adult and Minor this should be in Foundation. DONE
BE38 reworked the issue to say exactly one. Pete: There is a lot of follow through needed on the detail on this. For instance other cases that need to be decided based on the lifecycle of the LEI. When is this updated, reset and so on. Impact of merges of organizations and so on. Keep it open. David will review Pete's comment. Has been vetted with GLEIF. Operational adjustment may be needed.
Pete: Is there documentation on the meeting between David and the GLEIF person? Not documented. MB, DN and MA had a meeting with the leads of the GLEIF and covered this about 6 weeks ago. We have also covered the question in multiple conversations with Jeff Braswell, not recorded.
BE37 may be out of date given the refactoring of Corporation. Mark as resolved.
BE35 note to remove it now out of date, it's been removed and reinserted, and
Has a more detailed definition. Revolved
BE34 done see today's feedback. LEI for Juridical persons the current model covers this. There may be some updates, based on previous comments by Kristin Hofstein about possible changes to the scope of LEI as it relates to high net worth individuals, day traders etc. If and when there are such changes in LEI we will adjust the concept in FIBO.
BE33 restriction issue in Executives. Need to look at it. Not sure what the proposed resolution is Pete raised it. This was from when we were about to publish the Beta 1, this is the fix made in the RDF/OWL while Elisa made a separate change in the VOM which was what was published. Pete disagrees with Elisa's change, needs discusson
BE30 done I think
BE26 think this is done
BE17 would be a consequence of changes Mike made in Foundations in reaction to an issue raised by Pete, to change Percentage from a data type, to a class with a relationship to the thing it's a percentage of. The result of this is hat Elisa has found some interesting reasoning results and possibly a lint error. The issue is now in Foundation not BE.
Action: Dean and Elisa need to look at this and recommend a solution.
BE7 Affiliate review and reactors is on today's agenda. Should be in this release of BE. Was dependent on the legal forms and is also dependent on ownership and control. Is a good concept for testing BE because it reverses so many aspects of BE.
JIRA Summary: hope to close out quite a few of these Some are still unresolved.
To the model...David summarizes work done between him and Mike Bennett on how to create the best abstractions for kinds of entities in BE. This is a multi faceted taxonomy. (as you would expect for an ontology). We had restrictions without abstractions, following earlier set of abstractions without the restrictions. As we reviewed the restrictions that define each class of thing, we were able to focus in on specific abstractions that related to the precise restrictions.
Slide 6 of today's deck refers.
Business Entity is synonymous with Company, and covers the range of legal forms which are considered to constitute a company. There are restrictions to add on this set of things. Relates to the entity being both formed and administered under commercial law (including law covering charitable objectives). Includes sole proprietorships in a way that many other logical sets defined by legal form do not. There is work to do in unincorporated entities, and Foundations (the latter was left out of BE original submission as more work was required). Governmental things are also complex and were left out of the initial BE submission for the same reason these still need the full SME treatment. There is a distinction to be teased out between functional entities and corporations, statutory bodies or chartered bodies that are specifically government agencies, and these are in turn separate from the
sovereign, tribe, municipality etc. that are governed and on whole beheld debt is issues.
Discussion and review. Also needed something defined by the restriction of having (or of being able to have) an LEI. Previously known as LegalEntity, but since that term has 2 separate well attested meanings, we should not use it but label this abstraction directly as what it describes, which is LEI capable entity. So what was "LegalEntity" is now "LEI Capable Entity" and is defined as having exactly one LEI.
Does not include sole proprietorships. Includes governmental and supranational, and some trusts (not personal trusts for example). Further discussions in GLEIF will form possible changes which we would then add to this facet.
Formal Organization - The child of this was Formally Constituted Organization. That label was misread by some, as the intent was something constituted by means of a contract among its principals. So we renamed it to reflect that includes corporations, partnerships LLCs, trusts and so on. But not Sole Prop since they are not formed by an agreement with themselves. Certain governments may move into corporations but MB cautions that if we use the word Corporation in a more slippery way w will need to be careful. The word Corporation has many uses, not all of which correspond to the set of things that are a member of the class in FIBO called Corporation. As long a we keep clear on that we should be OK. For instance many of the governmental entities set up to do something, may have Corporation in their name (like the BBC). but these are not a member of the set of things defined by a contract among their principals (i.e. contractually constituted organization).
Main change reintroduction (removal) of Legal Person.
We have extended the abstractions n this slightly to broaden the meaning, but it relates to a clearly understood concept of fictional and real legal personhood. So now it is something with the capacity for liability, including the ability to be sued
(anyone you can collect the debt from even if thy are not capable of liability after all). Also it is further broadened by adding the right to sue someone else. So these 3 assertions are bundled together under Legal Person. Covers natural persons who are also legal person, and all artificial persons (juridical persons which are artificial legal persons. Previous there was ambiguity about some partnerships being juridical persons, given the joint and several liability of the general partners an their potential unlimited liability. What we have learned from the attorneys is that it is correct to define a general partnership as a legal person since it can also be named in a suit and its assets can also be gone after. In a General Partnership both the general partners (GPs) are fully labile, but the partnership is as well. So now, the litmus test for whether something is a juridical person, is if it can be named in a lawsuit. This means that unlike before, all forms of partnership are juridical persons (and therefore legal persons). There was s similar ambiguity about trusts, but less consensus among the attorneys. One of them is consulting a specialist in Trusts. There are also overrides in state law. So some of this stuff may not be 100% and there may be exceptions in many of these concepts which ay be discovered in law.
What we have are abstractions that will mainly be truthful from the perspective of a bank. Also where this behavior that will allow the court to pierce the corporate veil, we don't need to document that in FIBO, since these are exceptions set by the court, to the concepts we have in place in the model. Majority opinion is that trusts are always named in a lawsuit, along with the trustee. Originally we have trust as not a legal person because the liability fell to the trustee. However, in reality the liability does not fall to the Trustee as a person, but only to them as the party in the role of Trustee; their liability is somehow limited by this process. This means that the trust itself as an entity will have unlimited liability. David believes there may be another example here of a relative thing having specific properties and relations that are only named for the relative thing and not for the independent thing. That is, the Trustee is sued, not as a private but as "The Trustee" of that Trust, and are named as such in the lawsuit. MB note that the pattern limitations on relative things, are in no way applicable to additional properties stated about a relative thing, which fall outside of that pattern, Indeed, relative things exist only to have properties that are specific to them, and there should be no limit on those (they are part of the pattern we are defining as a set of modeling limitations). See MB note in response to Mike Uschold on upper ontology partitions and patterns, in the FLT wiki. Being able to assert additional properties of a relative thing, is something we need to be able to do, and examples like this are a good illustration of why.
There are already relationships in the published FBC ontology for Trusts, which are
relative to relative relationships. This also agrees with the previous EDM Council SME review work and with Elisa's experience as a trustee .
To Protege… In the current RDF OWL, David has kept Natural Person, it has no assertions but wants to get consensus of though group before removing it. Elisa: we need it in FBC. Registrar has the identity of a Natural Person, as does a Bank examiner and many other kinds of functional entity in FBC. Also the various officers of companies and so on, in Executives ontology. Also registrars and certain regulator roles. MB clarifies that this is not the deletion of a concept it is the renaming of a concept, and so any impact (which may still be considerable in the documentation) is only the impact of such a name change. The words Natural Person is more often used in reality to mean any human being. In the current published draft of BE we had used the label Natural Person to specifically mean a legally capable human being. Hence the name change.
EK Caution: when we write the resolution for the OMG to vote, we will have to state DAvid's rationale above, and to make it clear that this is a name change for a concept that continues to be in play. There are alternatives such as defining an equivalency.
The concept which is understood from the label "Natural Person" is different to what we previously thought. EK is sure that the original definition made it clear that this was a person that was legally capable.
Action: Pete reminds us of the importance of using synonyms. We should add a synonym of Natural Person to Legally Capable Person. All agree
MB: We should always use the synonym metadata and never use equivalent class, since we are doing n ontology not a vocabulary (classes are concepts not words). This is in the Legal Persons ontology. We must therefore ensure that if this is actions by creating a new class instead of renaming the old one, then we need to move all the metadata across. If we delete something in OWL we need to show the lineage. This would be required if we are eliminating a redundancy but not eliminating the concept (since the Legally Capable Person class already exists in Foundations). Also would update Person and include Natural Person as a synonym of Person. That is, Natural Person is a synonym for two separate concepts: a thing with human DNA, and a thing with human DNA and also legal personhood.
Action: MB to raise a new JIRA issue in Foundations to update the concept and put the synonyms and so on in there. DONE
At the same time David will do a JIRA issue in BE for the corresponding changes in that ontology. DONE
David shows the restrictions on LegalPerson. There is a technical discussion needed about being consistent across the different FIBOs, about when to do cardinality restrictions and when o use existential restrictions. MB notes that we should always use whichever restrictions reflect the semantics of the real work situation, this is not some decision that can be made on some technical basis ahead of modeling real things. There are not subtle differences, and nor are they technical decisions. We have already committed to using all of the OWL language so we need to be able use all the kinds of restrictions that OWL allows us.
EK: We are using the OWL language as our standard. As long as we use each of those kinds of restrictions properly then the semantics will be correct. This is not a technical matter and it is not trivial. Each kind of OWL restriction, and whether they are qualified or not, each has a very distinct meaning.
To clarify: allValuesFrom means the thing may only be one or more of those but could be none;
someValuesfrom means that all uses of that property for that restriction must include at least one f those things, but may also have others.
Cardinality is more specific as to numbers and the meanings if these are clear.
These three kinds of restrictions have entirely different semantics. David is looking for some standardized way of using restrictions that can be decided in advance without needing to understand the meanings. Everyone else disagrees. We need to be able to refer to and understand how to use the OWL standard. We do not need to come up with our own set of guidelines, OWL is it.
David still feels that there are changes to be made that would replace one kind of
restriction with another one, for reasons that are not clear. David would appreciate some guidance in the wiki or somewhere about when to use a cardinality restriction and when to use an existential restriction.
Cardinality restrictions are global, value restrictions are local. when you invoke them, if the application you are building uses the OWA, the results are different between use of value restrictions versus use of number restrictions. You also can't use number restrictions with transitive properties. There are serious differences in the semantics. If you are testing against databases, the result’s make no difference because you are working in the closed world. In the open world the differences are considerable and are important. Need to be able to say why a given restriction is written in the way that it is, with reference to a real use case (and to the real world stuff we are modeling). If we were only working in CWA things would be different. But FIBO is explicitly Open World
Action: Elisa, MB, MU, Dean in creating an explanation about when to use a cardinality restriction and when to use an existential restriction.
Pete Rivett: Use of a database does not imply "closed world"
David wants people who do not fully understand OWL, to nonetheless work with OWL within the FIBO ecosystem.
Juridical Person - this is a legal person which is artificial (and therefore is disjoint with Legally Capable Person. Then LCP in Foundations becomes a sub class of Legal Person, and the legally capable person is already (in FND) a union of emancipated minor and legally capable adult. This is in FND and there is no open issue to change this. This (Foundations) model does not cater for stateless persons. It makes it seem that a stateless person has no legal personhood. This might in fact be the case, but are we sure about this? For instance some stateless persons are given a passport for Switzerland, precisely so that they can exercise personhood.
David makes the case that we need to be able to say something about jurisdiction, since legal personhood is defined in terms of rights and obligations in a legal system.
Separately, when you enter into a contract the contract itself defines the jurisdiction in which it is struck, which is independent of the jurisdiction(s) of the parties to the contract.
Pete: We should have the relationship not to Jurisdiction only on Juridical person. How would we have legally capable person without reference to a jurisdiction? Can a stateless person be recognized as a legal person ? The stateless person can still be recognized as a legal person in any number of jurisdictions. The key thing here is that the property is 'is recognized in' which is distinct from the jurisdiction in which some juridical person is registered or (as a separate meaning again) constituted.
Checked the definition: looks OK except that it currently implies only one jurisdiction when we now see it could be many. With that change we are all happy.
Action: David will raise a JIRA issue with the before and after text for that definition we just changed.
LEI Capable Entity (formerly LegalEntity) we do need to show which entities are going to have LEIs. Is this 'capable 'or 'having' an LEI, or are we mingling these 2 concepts together? We are mingling them, as we are defining something that must have and also something that is capable of having an LEI and making them the same thing
PR: They are not the same thing There are many entities which would be technically
capable of having an LEI.
Review for next week:
See in Protege, the list of all things that are juridical persons, all things which are capable of having LEIs. Also please review whether LEI is must have or capable of having.
There is also the list of Business Entity (company) types.
Also the list of contractually constituted organizations.
Please review these lists for next week so we can wrap up legal forms.
The only potential major changes are in Governmental Entities, which were not in the original submission because we knew thy would need a considerable amount of work hence these were originally slated for inclusion in BE v 1.1. Some of these will functional, some will be statutory bodies, some might even be corporate .
- Elisa Kendall Dean and Elisa need to look at this and recommend a solution.
FND-29Getting issue details...
Dave Newman Pete reminds us of the importance of using synonyms. We should add a synonym of Natural Person to Legally Capable Person. All agree - BE-61Getting issue details... STATUS
- Mike Bennett MB to raise a new JIRA issue in Foundations to update the concept and put the synonyms and so on in there. - FND-30Getting issue details... STATUS
- Michael Uschold Elisa, MB, MU, Dean in creating an explanation about when to use a cardinality restriction and when to use an existential restriction. - INFRA-97Getting issue details... STATUS
- Dave Newman David will raise a JIRA issue with the before and after text for that definition we just changed.
BE-62Getting issue details...