From OntologPSMW

Jump to: navigation, search
[ ]
    (1)
Session Matthew West
Duration 1 hour
Date/Time 19 February 2020 17:00 GMT
9:00am PST/12:00pm EST
5:00pm GMT/6:00pm CET
Convener KenBaclawski
Track Why

Contents

Knowledge graphs, closely related to ontologies and semantic networks, have emerged in the last few years to be an important semantic technology and research area. As structured representations of semantic knowledge that are stored in a graph, KGs are lightweight versions of semantic networks that scale to massive datasets such as the entire World Wide Web. Industry has devoted a great deal of effort to the development of knowledge graphs, and they are now critical to the functions of intelligent virtual assistants such as Siri and Alexa. Some of the research communities where KGs are relevant are Ontologies, Big Data, Linked Data, Open Knowledge Network, Artificial Intelligence, Deep Learning, and many others.     (2A)

Agenda     (2B)

Presenter     (2B2)

Matthew joined Information Junction, as a director in 2008, where he delivers consultancy in information management, and is the Technical Lead for the UK Digital Twin programme. He gained a BSC and PhD in Chemical Engineering and joined Shell in 1978, initially as a Refinery Technologist, but from 1987 on the computing/business interface with a particular interest in Information Management and Data Modelling. He has been a key technical contributor to ISO 15926 "Lifecycle integration of process plant data including oil and gas production facilities" and IIDEAS (ISO 18876) - Integration of Industrial Data for Exchange, Access and Sharing. He also developed a data model of Shell's Global Downstream business.     (2B3)

Matthew was the Shell Visiting Professor in the Keyworth Institute at the University of Leeds. He is a Member of the Institution of Chemical Engineers, the British Computer Society, and he is a Chartered Engineer and a Chartered IT Professional. He is a member of the Ontolog Board of Trustees.     (2B4)

He is the author of "Developing High Quality Data Models" and was awarded an OBE for services to information management in the 2021 New Years Honours List (UK).     (2B5)

See also Dr. West's Ontolog Invited presentations:     (2B9)


Abstract: What is information for?     (2B14)

In business, information is used to support decisions. If information required for a decision is missing or inaccurate, this increases the risk of a mistake. So, to support a decision, information needs to be fit for purpose. This means that information management is a quality management process where information is the product.     (2B15)

But how do you know what your information requirements are? It turns out that asking people for their requirements give unreliable results. A better approach is to document your processes to the level where key decisions are identifiable. It is then possible to document the information requirements for those decisions.     (2B16)

Information has a lot of properties, but only some of them are critical for its use in supporting decisions. One of the hardest to achieve is consistency. If data is consistent then when it arrives from different sources, it can just be brought together and used immediately. This means that it uses the same data model and reference data (or if you prefer, they are knowledge graphs of the same ontology).     (2B17)

In the UK there is a project to develop a Digital Twin of national infrastructure. The aim is to establish a distributed Digital Twin of consistent data so that authorized users can construct queries across the Digital Twins so it is possible to answer questions like "Which Tower Blocks have the same type of cladding as Grenfell Tower?"     (2B18)

The Information Management Landscape sets out the information needed to support the critical properties of data and the information quality management process. Part of this is an integration architecture so that the distributed National Digital Twin can be virtualized so users see it as a single database, with only access to the data they are authorized to see.     (2B19)

  • General Discussion. All attendees are invited to participate in the discussion. To ask a question or make a contribution, please "raise your hand" on the Chat Room. The meeting will be recorded and the recording will be posted on this meeting page.     (2B21)

Conference Call Information     (2C)

Attendees     (2D)

Discussion     (2E)

[12:00] KenBaclawski: The slides are available at https://go.aws/2HdGBYr     (2E1)

[12:15] RaviSharma: Matthew - what is the meaning or definition of Digital Twin and what purpose does it serve?     (2E2)

[12:16] RaviSharma: Is it also to be used for disaster recovery and business continuity like a huge back up site with duplicated knowledge workers who keep governance alive despite disasters?     (2E3)

[12:21] Mike Bennett: Grenfell Tower would hardly be a simple query - as Matthew noted earlier, need to define what is meant by 'tower block' in order to query on towerblocks with a given kind of cladding.     (2E4)

[12:23] ToddSchneider: What are 'Physical Twin Processes'?     (2E5)

[12:24] TerryLongstreth: @Mike Bennett: & translate from UK to US: high-rise apartment building     (2E6)

[12:24] Mike Bennett: @Terry how high?     (2E7)

[12:25] ToddSchneider: Micheal, isn't the term 'towerblock' just a common term that refers to a building structure of a certain type for housing (people)?     (2E8)

[12:26] Mike Bennett: @Terry no. Many of the public housing estates in England are low-rise, 4 or 5 floors high.     (2E9)

[12:26] RaviSharma: how do you maintain sync     (2E10)

[12:26] Mike Bennett: The point is you need an ontology that sets out the characteristics in order to define a coherent category of thing to query against.     (2E11)

[12:26] TerryLongstreth: Mike; we have those too. They're usually (not always) called "garden apartments"     (2E12)

[12:27] Mike Bennett: @Terry the lexical space between UK and US is unlikely to map neatly to the semantic space. Why we need ontology not words.     (2E13)

[12:29] TerryLongstreth: Mike; I agree. My original comment was attempting to illustrate that. But the point should be made that people can't easily interpret ontologies. That's why TC37 (Terminology) struggles with the distinction between terminology and ontology     (2E14)

[12:29] Mike Bennett: @Terry or to each other of course. Even the way we measure number of floors is different.     (2E15)

[12:30] RaviSharma: so it is about how you integrate knowledge graph     (2E16)

[12:30] RaviSharma: thanks     (2E17)

[12:30] Mike Bennett: @Terry i am surprised to hear that. Lexical and semantic are clearly distinct. Only naive business people are expected to not know what ontology is and therefore confuse it with ontology.     (2E18)

[12:30] BobbinTeegarden: @Matthew The Messaging system carries both process and data? Is there a required structure for it?     (2E19)

[12:30] Mike Bennett: ^^ with terminology     (2E20)

[12:30] RaviSharma: would any simulation be a digital twin     (2E21)

[12:31] RaviSharma: foundation data model is ontology     (2E22)

[12:32] RaviSharma: buffer database is KG in slide 12?     (2E23)

[12:32] TerryLongstreth: @Mike: I agree, but either (or both) are used to derive semantics.     (2E24)

[12:35] Mike Bennett: So if I have a PID circuit made of analogue electronics, is that an analogue twin?     (2E25)

[12:35] RaviSharma: The solvers in packages for design for flow CFD or behavior e.g. stress bending moments etc are these digital twins of physical twins?     (2E26)

[12:37] RaviSharma: Ram - says it is a matter of details     (2E27)

[12:39] RaviSharma: Notes Ram says depth of detail - virtual dynamic design like CAD but Matthew is representing large whole city /     (2E28)

[12:40] RaviSharma: MW said authorization, virtualization are challenging.     (2E29)

[12:42] RaviSharma: Mike Richey talked about human capital     (2E30)

[12:43] RaviSharma: cognitive intelligent agents codification ML with ontology with business processes     (2E31)

[12:44] Leia Dickerson: Could you provide this slide in addition to your other slides? It is helpful.     (2E32)

[12:45] janet singer: If the digital twin is counterposed to a physical twin it seems to include a hope that the target can be pinned down by defining it as physical when something like a nation is not simply physical     (2E33)

[12:47] BobbinTeegarden: It looks like Business Process is driving the knowledge and information requirements, which is great! Isn't 'AI', the thinking part of your Digital Twin, mostly in the process?     (2E34)

[12:47] Mike Bennett: @Janet interesting point. I would have assumed that a DT can twin any 'real' thing, including those made from social constructs. Surprising if not.     (2E35)

[12:48] Leia Dickerson: @Janet -- I also like your point.     (2E36)

[12:51] janet singer: @Mike It can, but I think the complexities that previous data/model management efforts came up against are being hidden by an optimistic new metaphor     (2E37)

[12:51] Mike Bennett: This question (non physical DTs) also has implications for the ontological stance asserted by the foundational ontology / TLO that is used e.g. Realist v Conceptualist.     (2E38)

[12:51] ToddSchneider: It there any thought to apply the 'Digital Twin' notion to the information management problem (for integrating data/information from large numbers of organizations)?     (2E39)

[12:52] RaviSharma: janet - very complex even more complex than shell nuclear holocaust survival?     (2E40)

[12:52] RaviSharma: at nation level?     (2E41)

[12:58] RaviSharma: MW - you have referenced data as information element, what about knowledge level digital twin architectures, i guess similar to slide 24 and beyond?     (2E42)

[12:58] janet singer: Matthew says previous efforts have been stymied by new requirements for data models that come in at 18 months; not everyone updates due to cost     (2E43)

[12:58] BobbinTeegarden: @Matthew: won't your TLO morph depending on context/circumstance?     (2E44)

[13:01] janet singer: Need to use a top-level ontology so the models will be easily evolvable/extensible with new requirements     (2E45)

[13:01] janet singer: (per Matthew)     (2E46)

[13:02] Mark Underwood @knowlengr: This framing feels static, while Spark/Plink/Splunk etc demand dynamic models that are self-reorganizing even as they are "steadied" in upper ontologies     (2E47)

[13:03] janet singer: Matthew: Digital twin is just a way of talking; synonym for information     (2E48)

[13:03] Mark Underwood @knowlengr: Which is where KG's are sometimes introduced as part of the scaffolding     (2E49)

[13:04] janet singer: @Mark: What is this steadied metaphor?     (2E50)

[13:04] Mark Underwood @knowlengr: The lack of ontologies for basic IT infrastructure itself is a huge handicap     (2E51)

[13:04] BobbinTeegarden: @Janet is Digital Twin JUST a synonym for information, not also action?     (2E52)

[13:04] Mark Underwood @knowlengr: @Janet: more later     (2E53)

[13:05] Mike Bennett: Main point: digital twins require us to formally set out the scope and requirements and the appropriate TLO ontological stance to support that (Realist, 4D, Conceptualist etc.) - e.g. Matthew includes DTs for future things     (2E54)

[13:06] janet singer: @Bobbin That was Matthews language. I agree with point of your question.     (2E55)

[13:06] RaviSharma: Mike Bennett and MW had interesting discussion on BFO, use of ontology (foundational) TLO on DT of beyond physical constructs, etc. Recorded in more detail.     (2E56)

Resources     (2F)

Previous Meetings     (2G)


Next Meetings     (2H)