The document describes an ontology dojo activity to help design a website for visitors attending a conference in Dublin. Participants mapped out the transit ecosystem in Dublin by listing objects, mapping them on the wall, and sketching in predicates to show relationships. Objects included trains, buses, taxis, payment types, and organizers. Participants then shared their progress, challenges, and questions. The dojo helped provide a shared vocabulary and understanding of the complex Dublin transit system to inform website design.
4. Ontology and the Development Life
Cycle
Need ProductRequirements
Design/
Architecture
Engineering Development Test/
Validation
House
• 3 Bedroom
• 2 Bath
• Home Office
• 1 Story
Inspection
5. Ontology and the Development Life
Cycle
Need ProductRequirements
Design/
Architecture
Engineering Development Test/
Validation
Wine
Website
Inspection
• Wine Varieties
• Find Information
• Common language
to exchange
information
• XML
• RDF
• OWL
• NEO4j
• SPARQL
Concept
Map
Entity
Relationship
Diagram
6. What is Ontology?
• Philosophical study of being and reality
• “Ontology” has taken on two meanings:
• Classical Ontology (a study of philosophy focused on the nature of
being)
• Modern Applied Ontology (Using ontological theory to describe
things to facilitate the design of information systems)
Ontology is a deliberate method of characterization
9. From Classical to Modern Applied Ontology
• Describe things and their
relationships to other things
• Model and document things
• Facilitate finding things and
sharing information
Prove and describe
existence of things.
Classical Modern
10. Modern Ontologies
• Highest Level
• “Metaontology”
• Rosetta Stone
Upper Level
• Scoped by a topic
• “Unique” Language (Controlled Vocabulary)Domain
• Mix of Upper Level and Domain
• Usually very technicalHybrid
16. Domains
The purpose of _________________________is to
____________________________ so
that___________________________
17. Domains
The purpose of the ______________Ontology is to
____________________________ so
that___________________________
Michigan Wine
Show the relationships between wines varieties
users can discover the wine varieties they like
22. Concrete
• Does not have to be something you “touch”
• Concrete objects can take action
• Adheres to the laws of physics
• Much easier to communicate and describe to
others
Objects
23. Abstract
• May have to “invent” physics
• Harder to describe due to perspective
• May not be able to take action
• Often reuse terms and concepts we are
familiar with.
Objects
24. On the edges
• Virtual objects can be hard to categorize
• May be a matter of perspective
• Interaction between abstract and concrete
objects can be interesting
Citizens United v. FEC (2010)
Objects
31. IS A TYPE OF
Animal
Invertebrate
Insect
Crustacean
Vertebrate
Mammal
Reptile
Uses
• Taxonomies
• Card Sorting
• Affinity Diagrams
• Concept Map
Hypernym
Hyponym
Predicates
32. IS PART OF
Uses
• Site Map
• Content Inventory
• Venn Diagram
• Card Sort
• Concept Map
• Affinity Diagram
Meronym
Meronym
Meronym
Holonym
Predicates
33. IS PART OF
Example Part Type
(1) The handle is part of the mug. Physical
(2) The remote control is part of the stereo system. Collection
(3) The left half is your part of the cake. Portion
(4) The cutlery is part of the tableware. Set
(5) The contents of this bag is only part of what I
bought.
Unrelated Collection
(6) That area is part of the living room. Spatial
(7) The outermost points are part of the perimeter. Location
(8) The first act was the best part of the play. Temporal
Different Types of Parts
Predicates
37. • Try to use predefined/
established predicates
• May have to create abstract
objects
• Make a legend/directory of your
predicates
Is a Type of
Is a Part of
Is Like
Is not Like
Is a Type of
Is a Part of
Is Like
Is not Like
Purchases
Reviews
Creates
Predicates
Other Predicates
38. • Try to use predefined/
established predicates
• May have to create abstract
objects
• Make a legend/directory of your
predicates
Tammy Pez
Friend of
Predicates
Other Predicates
39. • Try to use predefined/
established predicates
• May have to create abstract
objects
• Make a legend/directory of your
predicates
Tammy PezTammy Pez
Friend-
ship
Predicates
Other Predicates
46. • Captures characteristics
• Almost endless data points can
be captured…know the scope of
your project
• Attributes/Metadata has evolved
tremendously over past 20 years
Name:
Height:
Weight:
Age:
Make: MG
Model: TD Midget
Body Style: Convertible
Color: Lt Brown
Era: Vintage
Make: MG
Model: TD Midget
Height: 53”
Length: 145”
Width: 59”
Make: MG
Model: TD Midget
Website: www.mg.co.uk
Description: The MG T series is a range
of body-on-
frame convertible sports
cars that were…
Video:
Attributes
50. Ontology IA/UX Comp Sci Business Library Science BOLO
Class • Class
• Group
• Collection
• Category
• Class
• Group
• Collection
• Class
• Group
• Collection
• Category
• Class
• Subject
• Schema
• Taxon
• Type
Classes
51. • When several object share
one or more non-
identifying attributes, they
may makeup a class
• Sales Manager is the class
in the example
• The more attributes that
match, the more specific
the class (North American
Sales Manager
NAME: Alan Moore
OCCUPATION: Sales Manager
LOCATION: North America
SALARY: $88,000
NAME: Sara Whatley
OCCUPATION: Sales Manager
LOCATION: South America
SALARY: $93,000
NAME: Torey Joliet
OCCUPATION: Sales Manager
LOCATION: North America
SALARY: $82,000
Classes
52. “To be or not to be”
“I think therefore I am”
“The definition of
‘is,’ is”
Class/Predicate/Attribute Pitfalls
53. Mammal
Is a type of
A Cat Maria
Mammal
Is a type ofClass
Instance
Is a
Class/Predicate/Attribute Pitfalls
57. Is a Part ofIs a Type of
HIERARCHIES (TAXONOMIES, SITE MAPS, AFFINITY DIAGRAMS)
Object
Object Object Object Object
ObjectObject
Object
Predicates
(Is a Type of)
(Is a Part of)
• Try to use one predicate
type per diagram
• Mix predicates vertically
Deliverables
58. PERSONAS (ORGANAS, SPEC SHEETS….)
SusanThe Administrator
Location: Apps:
iControl
NetView
HRBase
Basic Intranet
Admin Stats
~~~: xxxxxxx
~~~: xxxx
~~~: xxxxx
Description:
Lorem ipsum dolor sit
amet, consectetur
adipiscing elit. Nullam
suscipit lacus sit amet eros
vehicula sagittis. In in
sapien turpis. Donec risus
nisi, bibendum id tristique
non, rutrum vitae.
Availability:
Content
Knowledge:
Training:
Investment:
P-12.AB20
Deliverables
59. SusanThe Administrator
Location: Apps:
iControl
NetView
HRBase
Basic Intranet
Admin Stats
~~~: xxxxxxx
~~~: xxxx
~~~: xxxxx
Description:
Lorem ipsum dolor sit
amet, consectetur
adipiscing elit. Nullam
suscipit lacus sit amet eros
vehicula sagittis. In in
sapien turpis. Donec risus
nisi, bibendum id tristique
non, rutrum vitae.
Availability:
Content
Knowledge:
Training:
Investment:
P-12.AB20
CLASS GLOBAL METADATA
CLASS METADATA
ID
PERSONAS (ORGANAS, SPEC SHEETS….)
Deliverables
60. CONCEPT MAPS
Title (Domain)
Is a Part of
Is a Type of
Uses
Object
Object
Object
Object
Object
Predicate
Predicate
Predicate
Predicate
Predicate
Attributes
Attributes
Attributes
Attributes
Attributes
Class
Deliverables
61. Is a Part of
Is a Type of
Uses
System ToolsAdministrator
Anchorage
Office
iControl
XYZ Corp’s Intranet
HRBase
NetView
Personnel
(Persona)
(Taxonomy)
(UI Wireframe)
P-12.AB20
CONCEPT MAPS—TYING IT ALL TOGETHER
Deliverables
65. High Fidelity
• Capture Metadata/Predicates/Classes
• Actionable ( Development Ready) Designs
• Record Keeping
Mind Mapping
CAD
Tools and Technologies
66. • Extensible Markup Language (XML)
• Unified Modeling Language (UML)
• Ontological Web Language (OWL)
• Resource Description Framework (RDF)
Tools and Technologies
67. Goal:
Complete a series of activities that simulate ontological thinking and
the creation of an ontology to solve a complex IA problem.
Problem we are solving:
We are building a website or application for a diverse group of
conference visitors to Dublin in 2018, and need to make a wide variety
of transportation options, and their interplay, clear
Ontological Thinking Activity
70. Mapping out a model of the complex
system will help us understand how to
best situate and plan our website within
the already complex Dublin Transit
ecosystem.
Mapping out this ecosystem will also allow
us to have a shared vocabulary around
stakeholders, users, inputs, outputs,
integrations, and relationships between
these entities, leading to more clear
shared communication on project goals
and complexities.
Ecosystem Model
73. Objects are nodes in a complex system.
Think of any noun you might use when
talking about transit in Dublin. They can
be:
• People / Roles
• Groups
• Organizations & Agencies
• Places
• Artifacts
• Payment types
Pro-Tip: group all like items on the same color
Post – it – ie: all people are blue, all systems
green.
Objects Rail
Car Sharing
Organizer
Attendee
Aircoach
Taxi
Bus
Luas
Trains
Leap Card
Permit
MyTaxi
74. • Looking at your Wikipedia articles, pull out:
• Objects (nouns)
• Classes (categories you need to call out)
• If time - Attributes (Adjectives)
• Generate other objects you know to be true but are not in the Wikipedia article (see the article
co-chairs put together, personal experience)
• Write on your post its – remember, ideally each class should have it’s own color
• Ren & Jess walk around and answer questions
• 15 minutes to do this
Finding your Objects
78. Predicates are typically verbs and
prepositions that show relationships
between objects.
Examples include:
• Is a part of
• Is a type of
• Is like
• Serves
• Purchases
Remember: Keep your predicates
streamlined, uniform, and minimal.
Predicates
81. It’s ok for your
Predicates to
be messy
at this stage
82. Group sharing and conversation:
1. Describe your progress & approach
2. Challenges
3. Open Questions
4. Reflections
Sharing
83. Ready to go out and
CLASS
First Round of Ontology Dojo
84. • What do you do with it once you're done?
• Solving the initial ask… how does this help us actually make the website now?
• Deriving sitemaps, taxonomies, user journeys from this ontology
• How could you start to do some organization with this? Sitemaps? Would different table have
different sitemaps? Why?
• Methods of output and application
• XML vs. UML vs. OWL vs. RDF - would be nice to see a chart comparing these
• Tools of the Trade
• CMAP
• Excel
• Other Software -(ie: Visio, Omnigraffle)
• Taxobank
Now What?
85. • Books, Blogs, Training, Events
• Carrie Hane & Mike Atherton: Designing Connected Content
• Carrie Hane – Connected Content @IAS 2018
• Kat King - Personal Ontology Maps @IAS 2018
• Ontology Worksheet
• https://ubwp.buffalo.edu/ncor/training-and-education/
• Stanford on Parthood
• Ontology
• Upper Ontology
• How to follow up with questions
• Ren: @info_do
• Jess: jess@rs21.io
Additional Resources
The description of a subject (The Ontology of…)
A methodology to incorporate into other practices
A set of tools to describe things and relationships
Want to explain that the broader the scope the more you want to borrow.
Concrete vs abstract from Sarte’s being and nothingness
Relationship – Organization i.e. how the object is ordered and arrayed with an organization of objects within a domain
AS FROM THE PERSPECTIVE OF THE IA (ABSTRACT FOCUSES ON VIRTUAL/CYBER)
Physics: There can only be one, can’t be in two places at once
Physics: There can only be one, can’t be in two places at once
Physics: There can only be one, can’t be in two places at once
Want to explain that the broader the scope the more you want to borrow.