From OntologPSMW

Jump to: navigation, search
[ ]
    (1)
Session The Ontological Landscape
Duration 1 hour
Date/Time 10 Feb 2021 17:00 GMT
9:00am PST/12:00pm EST
5:00pm GMT/6:00pm CET
Convener Ravi Sharma
Track A and D

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)

Conference Call Information     (2C)

Attendees     (2D)

Discussion     (2E)

[11:43] Ken Baclawski:     (2E1)

[12:10] Mark Underwood: Low urgency remark: Remove copyright. We want to socialize assiduously.     (2E4)

[12:10] David Eddy: any room yet for "un-natural language" or "messy language"... the sort of stuff we write in software?     (2E5)

[12:11] David Eddy: Or is that still non-existent?     (2E6)

[12:14] DrRaviSharma: David - what is the distinction between software language and ML?     (2E7)

[12:15] Ram D. Sriram: The URL for the Ontology, Taxonomy, Folksonomy: Understanding the Distinctions summit is https://ontologforum.org/index.php/OntologySummit2007     (2E8)

[12:15] David Eddy: @Ravi... (by no means am I schooled in ML)... but as far as I know, the whole arena of AI/ML/NLP, etc totally ignores the language used in working software.     (2E9)

[12:15] EmilioSanfilippo: Do you characterize/recognize a distinction between foundational ontologies and upper-level ontologies?     (2E10)

[12:16] Andrea Westerinen: If a foundational ontology is based on specific philosophical concepts, does that make it "common across all domains" or more specific to a philosophy domain?     (2E11)

[12:16] Andrea Westerinen: That might indicate whether it is useful or not in an application.     (2E12)

[12:17] DrRaviSharma: Todd - consistent semantics take us toward harmonization?     (2E13)

[12:20] David Eddy: ....ok... slide 13 certainly seems to eliminate any interest in the other side of the NLP coin... "messy language" used in working software     (2E14)

[12:22] DrRaviSharma: Todd- Would controlled or computational version of NL help improve the Human Machine interaction towards ontology development?     (2E15)

[12:22] David Eddy: In the world of data, only the data is of interest... the "tools" used to manufacture data of no interest     (2E16)

[12:27] DrRaviSharma: Todd- Reusability of an ontology indicates better harmonization and or better interoperability or just better extensibility?     (2E17)

[12:29] Andrea Westerinen: OWL2 + SWRL or a rule language is quite expressive. Many OWL tools also include rule support.     (2E18)

[12:31] DrRaviSharma: Todd - slide 14 diagram is valuable as it identifies many aspects and is a good view of landscape related value. Thanks     (2E19)

[12:32] Leia D.: Who are the people suitable for leading ontology development, profession wise?     (2E20)

[12:37] Andrea Westerinen: @leia Here is a job description that I wrote:     (2E21)

[12:37] Andrea Westerinen:     (2E22)

  • (New) Job Description:     (2E23)
    • Working with subject matter and natural language processing experts to develop graph-based semantic models (including taxonomies and ontologies) which capture knowledge about diverse real-world scenarios     (2E23A)
    • Developing tooling to support capture of domain concepts and to allow model/ontology extension by subject matter experts     (2E23B)
    • Creating tooling to effectively query, align and integrate models/ontologies     (2E23C)
    • Creating documentation, training materials and tooling to enable understanding and effective use of the models     (2E23D)
  • (Updated) General skills:     (2E24)
    • Senior developer with experience on a range of products (and an ability to draw analogies and see commonalities between the products)     (2E24A)
    • Ability to discern use cases and create abstractions from basic problem statements     (2E24B)
    • The ability to solve a single problem with a specific solution is relatively easy, but the ability to create a generic solution to a set of related problems is much more valuable and difficult     (2E24C)
    • Understanding of systems thinking and an ability to define "systems", determining boundaries and components     (2E24D)
      • For example, looking at a few trees and abstracting the forest - or - looking at a forest and determining what is important about the trees     (2E24D1)
    • Strong interpersonal and communication skills     (2E24E)
      • The job is really all about understanding requirements for the ontology/model, learning about the domain and interacting with subject matter experts     (2E24E1)
    • Ability and eagerness to learn (about new problems, new areas of work, specific details that need abstraction, ...)     (2E24F)
  • (Updated) Specific experience:     (2E25)
    • Minimum of 2 years experience in creating and/or editing graph-based data models and/or ontologies (experience both creating and evolving preferred)     (2E25A)
    • Experience using RDF and/or OWL     (2E25B)
    • Experience using a graph query language (either GraphQL or SPARQL; experience with both preferred)     (2E25C)
    • Minimum of 5 years programming experience (Java and JavaScript preferred)     (2E25D)
    • Some natural language processing experience, for example working with entity recognition or concept extraction / topic modeling     (2E25E)
    • Ability to communicate complex ideas clearly and persuasively     (2E25F)

[12:38] Cas Miles: http://www.meteck.org/ <-- Maria Keet     (2E27)

[12:38] Leia D.: @Todd -- I've taken the EDM Council's course. Also have some books.     (2E28)

[12:39] LaureVieu: IAOA is planning a new series of on-line short courses. Keep tuned!     (2E29)

[12:41] David Eddy: Useful Getaneh Alemu & Brett Stevens: "An Emergent Theory of Digital Library Metadata: Enrich then Filter" Huge chasm between world of librarianship & the software world     (2E30)

[12:41] Leia D.: @LaureVieu-- Thank you!!!     (2E31)

[12:42] DrRaviSharma: @Dave Hay - please provide URL of your LOC talk.     (2E32)

[12:44] AlexShkotin: ??Data Model Patterns are useful. If you go to loc.gov and look up Data Model Patterns     (2E33)

[12:45] Gary Berg-Cross: Data Model Patterns are useful. If you go to loc.gov and look up Data Model Patterns     (2E34)

[12:45] AlexShkotin: I have just copy-pasted from Queue.     (2E35)

[12:45] Leia D.: @Andrea -- Thank you for your thoughts.     (2E36)

[12:47] Leia D.: @Andrea -- also have experience problem solving through research (literature searches) as well as some data management...     (2E37)

[12:48] Andrea Westerinen: @leia And some knowledge engineering     (2E38)

[12:48] Leia D.: @Andrea. Thank you.     (2E39)

[12:50] Dave Hay: Leia D, Can I refer you also to my first book, "Data Model Patterns: Conventions of Thought". It is at a level of abstraction that has broad applicability, but concrete enough to be understood. In my later book, "Enterprise Models: Describing the world" specifically addresses the different levels of abstraction that Todd describes here.     (2E40)

[12:51] Leia D.: @Dave -- Thank you.     (2E41)

[12:55] David Eddy: @Ravi... I thought it has been firmly established that ontology work is not related to software development/understanding     (2E42)

[12:58] Dave Hay: The problem with subject matter experts is that their world is complex. It is the objective of the ontologist to redescribe their world in simpler terms.     (2E43)

[12:58] Leia D.: Thank you all. I have to go to another meeting.     (2E44)

[13:00] Andrea Westerinen: Thanks... have to leave also.     (2E45)

[13:02] Janet Singer: @Andrea, @Todd: Seems the problem arises when foundational ontologies are treated as getting their authority from philosophy rather than being evaluated as engineering artifacts. As the latter they need to demonstrate fitness for purpose (making distinctions common across domains, facilitating clear thinking and communication in conceptual modeling, etc.)     (2E46)

[13:05] DrRaviSharma: Thanks to all on behalf of Todd and me as well as Ken for an interesting session.     (2E47)

[13:06] Janet Singer: Actually Todd has laid out the issues and criteria. Its just a problem when foundational ontologies are treated as exiting outside the evaluation criteria for all ontologies     (2E48)

[13:07] Cas Miles: Thank you all!     (2E49)

Resources     (2F)

Previous Meetings     (2G)


Next Meetings     (2H)