A Learning Trajectory for Building Ontologies

1
A Learning Trajectory for Building Ontologies
The case of GEON, and building ontologies

• David Ribes / Geoffrey Bowker
• University of California San Diego / Santa Clara
  University
• Sociology / Science Studies

2
Outline

• What is this research about?
• GEON, Ontology Data Interoperability
• A Learning Curve for how to ontology
• Translation What is it to formalize knowledge
  into an ontology?

3
What is this research about?

• GEON (www.geongrid.org)
• cyberinfrastructure for the geo-sciences
• systems computing resources and Grid development
• knowledge representation ontologies, workflows
  c.
• tools mapping, visualization

4
What is this research about?

• GEON (www.geongrid.org)
• cyberinfrastructure for the geo-sciences
• systems computing resources and Grid development
• knowledge representation ontologies, workflows
  c.
• tools mapping, visualization
• Data Interoperability Strategy Two Phased
• 1- Systematic Ontology Construction
• -supporting current day data-practices while
  enabling interoperability
• 2- Seamless migration to community standards

5
DATA!
6
Heterogeneous Data integration
DATA!

• Requires advanced metadata and processing
• Attributes must be semantically typed
• Collection protocols must be known
• Units and measurement scale must be known

7
Why develop an ontology?
DATA!

• To make domain assumptions explicit
• Easier to change domain assumptions
• Easier to understand, update, and integrate
  legacy data
• ? data integration
• To separate domain knowledge from operational
  knowledge
• Re-use domain and operational knowledge
  separately
• A community reference for applications
• To share a consistent understanding of what
  information means.

Carole Goble, Nigel Shadbolt, Ontologies and the
Grid Tutorial
8
Why develop an ontology?
DATA!

• To make domain assumptions explicit
• Easier to change domain assumptions
• Easier to understand, update, and integrate
  legacy data
• ? data integration
• To separate domain knowledge from operational
  knowledge
• Re-use domain and operational knowledge
  separately
• A community reference for applications
• To share a consistent understanding of what
  information means.

Carole Goble, Nigel Shadbolt, Ontologies and the
Grid Tutorial
9
A Process Model of Ontology Building
10
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
Community Acceptance And Usage
Ontology
11
A Process Model of Ontology Building
Knowledge Community
12
A Process Model of Ontology Building
IT
Knowledge Community
13
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
14
(No Transcript)
15
Elevation Latitude/Longitude UTM USGS 7.5 Quad
Location
Rocks
Metamorphic Facies
Fit into an explicitly defined
Name
Defined by
isograds
Physical Properties
Mineral Assemblage
Metamorphic Grade
Fabric/Texture
Outcrop
Hand specimen
Mineral
Minerals
Inclusions
Metamorphic reactions
Thin sections
Fluid
Structure
Inclusions
P-T-X Space
Individual mineral ages
Abundance (Mode)
Composition/Chemistry
Isotopes
P-T-t paths
Classification
P-T-Xfluid estimates
Major oxides
Trace Elements
M rxns
Terrane Recognition
Metamorphic History
Classification
W.C.
16
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
17
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
Ontology
18
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
Community Acceptance And Usage
Ontology
19
A Process Model of Ontology Building
Knowledge Acquisition
IT
Knowledge Community
Community Acceptance And Usage
Ontology
20
A Process Model of Ontology Building
Knowledge Community
21
Learning Curve
Knowledge Acquisition
Community Acceptance And Usage
Ontology
22
Learning Trajectory

• Enrollment of
• scientists
• Practice / Formalization
• 3. Enrollment of
• community

Knowledge Acquisition
Community Acceptance And Usage
Ontology
23
1- Enrollment of scientists

• Enrollment is the initial process of engaging
  domain scientists in the process of ontology
  building
• What is an ontology?
• The technical is opaque and irrelevant to
  scientists
• Why should scientists invest their time?
• Few science results or career rewards
• Why should they chose this method?
• Federation and community standards
• IT Team learns Domain Data Politics

24
2- Practical Learning - Formalization
25
2- Practical Learning - Formalization

• In the actual work of building ontologies a
  specialist typically finds that domain knowledges
  are not readily prepared for ontology building.
  Written sources such as textbooks and technical
  treatises are often not precise enough for
  transformation into description logics there may
  be competing accounts of the same phenomena,
  overlapping taxonomies and standards, or outright
  contradictions. Similarly in consulting
  authorities in a domain, a programmer may find
  that these experts are not immediately able to
  state domain knowledge in the terms necessary for
  ontology building

26
2- Practical Learning - Formalization

• Why?

27
Knowledge in the wild

• has a tacit dimension not immediately available
  to discursive expression
• is embodied it is known as an action in
  practice as well as being a formal set of
  propositions
• is immanent it is developed to respond to
  certain problem sets and is not necessarily
  immediately generalizable beyond these
• is heuristic knowledge may serve to help
  understand a phenomena, even when practioners do
  not believe the knowledge faithfully represents
  reality (all models wrong)
• is nebulous - terms and concepts familiar to
  domain experts may seem fuzzy, incomplete or
  vague to outsiders
• is multiple each sub-domain has its own set of
  methods, tools, heuristics and criteria for truth
  
• is contentious knowledge is often the focus of
  debate, struggle and contestation science is
  both a repository for knowledge but also a site
  for its development

28

• Domesticating Wild Knowledge can be a politically
  charged act for scientists!

29
Observed Strategies for Representing Wild
Knowledge

• i- scale up deal with the issue at a higher
  conceptual level where the domain has established
  a stronger consensus
• ii- pragmatics encourage the domain experts to
  form a working consensus in order to continue the
  process
• iii- raincheck leave the problem aside for a
  later time, experts may be able to resolve the
  issue with a review of evidence, consultation
  with experts or the production of new evidence
• iv- represent the uncertainty technical
  solutions within ontologies work permit encoding
  multiple knowledges, disagreement, uncertainties,
  ambiguities or ambivalences

30
3- Enrolling the Community
31
3- Enrolling the Community

• the frontstage work of formalizing knowledge,
  forming temporary pragmatic consensuses, or
  representing uncertainties, must be coupled with
  the backstage work of securing consent,
  building alliances, and holding standards in
  place within a community

32
A Process Model of Ontology Building
Community Based Ontology Workshops
Knowledge Community
33
3- Enrolling the Community

• It has become understood that
• In returning to the community, a contentious
  ontology can mean that knowledge developed
  derived from the ontology could itself be
  contentious.

34
Further Research Comparative Interoperability
Project

• interoperability.ucsd.edu
• Community driven standard formation
• Long Term Ecological Research (LTER)
• - Ecological Metadata Language (EML)
• Participatory Design
• -Ocean Informatics at Scripps Institute for
  Oceanography

35
Fin