From OntologPSMW

Jump to: navigation, search
[ ]
    (1)
Session The Ontological Landscape
Duration 1 hour
Date/Time 03 Mar 2021 17:00 GMT
9:00am PST/12:00pm EST
5:00pm GMT/6:00pm CET
Convener Todd Schneider
Track A

Contents

Ontology Summit 2021 The Ontological Landscape     (2)

Ontologies are a rich and versatile construct. They can be extracted, learned, modularized, interrelated, transformed, analyzed, and harmonized as well as developed in a formal process. This summit will explore the many kinds of ontologies and how they can be manipulated. The goal is to acquaint both current and potential users of ontologies with the possibilities for how ontologies could be used for solving problems.     (2A)

Agenda     (2B)

Ontologies are produced for a number of different reasons, while the use of modeling languages such as OWL bring certain assumptions and constraints to the effort. This has implications for ontology reusability. One particular distinction is whether the ontology represents real things in the world (of conceptualizations of those real things) or is really a model for data about those real things. For example if the formal meaning of something is based on some legal or social construct, there may not be data explicitly representing that construct. Or there may be data framed according to a typing system other than that supported by RDF and OWL.     (2B3)

In this session Mike Bennett will present a proposed 'landscape' of kinds of ontology, distinguished according to their purpose (semantic applications use cases, data integration, AI, business knowledge representation and so on). This will be followed by an open discussion in which we hope to hear and share as many perspectives as possible on how to characterize what kinds of ontology there are out there and what are the ideal ontology characteristics for these and other purposes.     (2B4)

Conference Call Information     (2C)

Attendees     (2D)

Discussion     (2E)

[12:07] RaviSharma: Mike what would it require to mature to information model rather than data model?     (2E1)

[12:07] RaviSharma: Actually things and relations and their transformations to information of value is the goal - how can we achieve it?     (2E2)

[12:09] RaviSharma: mike - why don't we call these information models?     (2E3)

[12:11] TerryLongstreth: @Ravi: are we restricting ontology to information systems? Can't the Concept(ual) Model describe business operations and requirements(independent of an electronic information system)?     (2E4)

[12:13] RaviSharma: mike- business and operational data types - they may not be same type as ontology concept to implementation operational ontology may be different data type combinations?     (2E5)

[12:13] ToddSchneider: Is there an assumption that 'data' exists a priori to the use of an (implementation of) an ontology?     (2E6)

[12:16] ToddSchneider: What is 'data'?     (2E7)

[12:17] RaviSharma: Mike - like physical data representing nature's models (physics) both are real as they are related by how truly they represent nature.     (2E8)

[12:18] RaviSharma: Most living items even enterprises are living and transforming things?     (2E9)

[12:20] RaviSharma: Mike - like people, loan is about say an object say home and loan has life origin to disposition     (2E10)

[12:22] RaviSharma: Mike - the last thing you imply could be that home is destroyed in fire but loan does not disappear but transforms to remaining liability or insurance or NPA?     (2E11)

[12:23] RaviSharma: Mike - seems like you are linking truthmakers to digital fingerprints but truthmakers may transform differently?     (2E12)

[12:25] Mark Underwood: I was about to comment on blockchain / distributed ledger surrogates, but Mike beat me to it     (2E13)

[12:27] RaviSharma: Mike- truthmaker when it transforms what happens, for example CDOs obliterated the fine structure and time diluted the credit history but both not linked? so how truthmaker and digital fingerprint diverge from each other in reality / or truth?     (2E14)

[12:29] Robert Rovetto: Some input re:truth-makers. We should take care with employment of the concept of truth-making, and be clear what's intended. Some context: it has a long history in philosophy, where we various accounts of what it is, e.g., as necessity, grounding, entailment, etc. So mixing it with computer science centric goals or concepts may vary widely.     (2E15)

[12:32] RaviSharma: Mike- concept to data ontologies at concept level it appeared that your were using TLO? for part of the graph?     (2E16)

[12:33] Pete Rivett: How would you associate data surrogates/fingerprints to classes in an ontology? Has someone proposed annotations for that?     (2E17)

[12:35] Mark Underwood: Comment: there are interesting data deltas in operationalizing privacy, increasingly a domain-specific enterprise drifting away from information technology     (2E18)

[12:40] RaviSharma: Mike - But you can use CL or FOL for Conceptual Ontology?     (2E19)

[12:40] Robert Rovetto: Re: not right with respect to physics, because of the abstract assumptions. We spoke about this is past IAOA meetings, but care is needed when using 'realism' and 'conceptualism' since there are various flavors     (2E20)

[12:41] RaviSharma: Mike - how does multiple inheritance work in ontologies?     (2E21)

[12:41] Robert Rovetto: Re: categorical name for abstract ontology...I recommend using a more generic term rather than 'tlo' because 'tlo' assumes a certain architecture that does not necessarily apply to all ontologies, let alone the most abstract or generic ones     (2E22)

[12:42] ToddSchneider: Ravi, for Track A, 'Ontology Landscape', on 10 March 2021 Claudio Masolo will be providing a presentation. And then on 21 April 2021 Michael Uschold will be providing a presentation.     (2E23)

[12:43] RaviSharma: how do you mature and measure what you just described? Levels?     (2E24)

[12:50] BobbinTeegarden: Doesn't TLO just refer to the top of the current domain of discussion? I.E. is a term related to the current context?     (2E25)

[12:50] Gary Berg-Cross: Digital data is a type of data allowing a class of processing.     (2E26)

[12:52] TerryLongstreth: The truth maker may be an oracle that cannot be captured in an ontology     (2E27)

[12:52] janet singer: Relying on non-randomness to define information raises the question of what randomness is...     (2E28)

[12:55] Robert Rovetto: @Bobbin: in the context of this presentation it was used in the sense of the (arguably) most abstract ontologies. So it would not be related to a given context, but a generic model providing a generic characterization (of which there are many) of the context. You raised and phrased a good question, echoing what i suggested.     (2E29)

[12:59] Robert Rovetto: will OntoClean help with answering those questions? (what are the N&S...)     (2E30)

[12:59] RaviSharma: mike - kindly find some moments to review our Questions on this chat and hopefully we can exchange emails.     (2E31)

[13:00] Gary Berg-Cross: You might formalize an "identity" relation to handle this issue.     (2E32)

[13:00] janet singer: All of these questions of defining terms in the ontology landscape (in this presentation, Todd's earlier one, and the issues Rob raises above, etc.) keep coming up against the need for clear ontological thinking on the meta or supra level covering the process of people developing ontologies as engineered artifacts.     (2E33)

Resources     (2F)

Previous Meetings     (2G)


Next Meetings     (2H)