Ontology Engineering

1
Ontology Engineering Ontology Acquisition

• Primarily Ontology Acquisition

2
Building Ontologies

• No field of Ontological Engineering equivalent to
  Knowledge or Software Engineering
• No standard methodologies for building
  ontologies
• Such a methodology would include
• a set of stages that occur when building
  ontologies
• guidelines and principles to assist in the
  different stages
• an ontology life-cycle which indicates the
  relationships among stages.
• Gruber's guidelines for constructing ontologies
  are well known.

3
The Development Lifecycle

• Two kinds of complementary methodologies emerged
• Stage-based, e.g. TOVE Uschold96
• Iterative evolving prototypes, e.g. MethOntology
  Gomez Perez94.
• Most have TWO stages
• Informal stage
• ontology is sketched out using either natural
  language descriptions or some diagram technique
• Formal stage
• ontology is encoded in a formal knowledge
  representation language, that is machine
  computable
• An ontology should ideally be communicated to
  people and unambiguously interpreted by software
• the informal representation helps the former
• the formal representation helps the latter.

4
A Provisional Methodology

• A skeletal methodology and life-cycle for
  building ontologies
• Inspired by the software engineering V-process
  model
• The overall process moves through a life-cycle.

The left side charts the processes in building
an ontology
The right side charts the guidelines, principles
and evaluation used to quality assure the
ontology
5
Methodology
Ontology in Use
Evaluation coverage, verification, granularity
Identify purpose and scope
Knowledge acquisition
Maintenance
User Model
Conceptualisation Principles commitment,
conciseness, clarity, extensibility, coherency
Conceptualisation
Integrating existing ontologies
Conceptualisation Model
Ontology Learning
Encoding/Representation principles encoding
bias, consistency, house styles and standards,
reasoning system exploitation
Encoding
Representation
Implementation Model
6
An Ontology Building Life-cycle
Identify purpose and scope
Consistency Checking
Knowledge acquisition
Building
Language and representation
Conceptualisation
Integrating existing ontologies
Available development tools
Encoding
Ontology Learning
Evaluation
7
Questions

• How do we obtain our conceptualisation?
• The role of texts, experts and other sources
• How do we derive conceptualisation from texts etc
• How do we cope with tacit conceptualisations?
• How do we use models with the expert?
• How do we validate the conceptualisation?

8
Knowledge Acquisition

• The process of capturing knowledge including
  various forms of conceptualisation from whatever
  source including experts, documents, manuals,
  case studies etc.
• Knowledge Elicitation
• techniques that are used to acquire knowledge
  direct from human experts
• Machine Learning
• use of AI pattern recognition methods to infer
  patterns from sets of examples

9
Problems of Knowledge Elicitation

• Techniques
• Limited range
• Ignorance
• Experts
• poor appreciation of different types
• ignorance
• Expertise
• poor appreciation of different types
• ignorance
• need to organise knowledge into higher level
  units

10
First Steps -Initial Understanding of the Domain

• Problem Description
• List knowledge resources (verify that knowledge
  really exists)
• Experts, Technical Authorities
• Text Books, Training Material
• Manuals and Procedures
• Databases and Case Histories
• Produce domain yellow pages
• Establish performance metrics
• Initial task environment analysis

11
Document and Text Analysis

• Look at the structure
• how material is organised into topics and
  sub-topics
• Content analysis
• Extract major linguistic categories
• nouns - objects and concepts
• verbs - relations
• modifiers - properties and values
• connectives - rules and links
• Use Intermediate representations
• Pseudo production rules
• Small concept networks and hierarchies

12
Problems of Document andText Analysis

• Documents and texts are written for specific
  purposes that may not reveal real knowledge or
  explicit concepualisations
• Duty logs and rostas
• Teaching texts
• All textual analysis is a form of content
  analysis - the interpreter may or may not be
  imputing the correct conceptualisation
• Difficult to reconstruct the context need to
  capture acquisition and design rationales

13
Types of Expert

• The Academic
• Values logical consistency
• The Professional
• Solutions that work in the context of information
  overload
• The Samurai
• Pure Performance
• State of knowledge varies
• Required solutions vary

14
Session Plan

• The importance of an acquisition plan
• A detailed agenda of what is to be covered during
  a KA session.
• Should include
• an introduction describing the objectives
• description of the techniques to be used
• questions to be asked (if required)
• timings
• Should be sent to the expert at least one day in
  advance of the session

15
KA Techniques

• Methods that help acquire and validate knowledge
  from an expert during a KA session.
• Three important types
• natural techniques
• contrived techniques
• modelling and mediating representation techniques

16
KA Typology
17
KA Techniques1

• Natural Techniques

18
Natural Techniques

• KA techniques that involve the expert performing
  tasks they would normally do as part of their
  job.
• Variations
• Interviews
• Observational techniques
• (Group meetings)
• (Questionnaires)

19
Interviews

• KA technique in which the knowledge engineer asks
  questions of the expert or end user.
• Essential method for acquiring explicit
  conceptualisations and knowledge, but poor for
  tacit knowledge.
• Variations
• Unstructured interview
• Semi-structured interview
• Structured interview

20
Unstructured Interview

• An interview in which the knowledge engineer has
  no pre-defined questions.
• Basically a chat to find out broad aspects of the
  experts knowledge.
• An aid to designing a KA session plan.

21
Semi-structured Interview

• An interview in which pre-prepared questions are
  used to focus and scope what is covered
• Also involves unprepared supplementary questions
  for clarification and probing.
• Questions should be
• designed carefully
• sent to the expert beforehand
• asked verbatim (read-out as written)
• include timings
• The recommended interview technique at the start
  of most KA projects.

22
Structured Interview

• An interview in which the knowledge engineer
  follows a pre-defined set of structured questions
  but can ask no supplementary questions.
• Often involves filling-in a matrix or generic
  headings.

23
KA Techniques2

• Contrived Techniques

24
Contrived Techniques

• KA techniques that involve the expert performing
  tasks they would not normally do as part of their
  job.
• Most of these techniques come from psychology
• Useful for capturing tacit knowledge, excellent
  for conceptualisations.
• Important types
• card sorting
• three card trick
• repertory grid technique
• constrained tasks
• 20-questions
• commentating
• teach back

25
Card Sorting

• KA technique in which a collection of concepts
  (or other knowledge objects) are written on
  separate cards and sorted into piles by an expert
  in order to elicit classes based on attributes.
• Also enables significant elicitation of
  properties and dimensions
• Used to capture concept knowledge and tacit
  knowledge
• Use in conjunction with triadic method
• Can also sort objects or pictures instead of cards

26
Triadic Elicitation Method

• KA technique used to capture the way in which an
  expert views the concepts in a domain.
• Involves presenting three random concepts and
  asking in what way two of them are similar but
  different from the other one.
• Answer will give an attribute.
• A good way of acquiring tacit knowledge.

27
Repertory Grid technique

• KA technique used for a number of purposes
• to elicit attributes for a set of concepts
• to rate concepts against attributes using a
  numerical scale
• uses statistical analysis to arrange and group
  similar concepts and attributes
• A useful way of capturing concept knowledge and
  tacit knowledge
• Requires special software (PC-PACK)

28
Repertory Grid Example
29
Repertory Grids -Demonstration usingPC-PACK
Laddering Tool
30
Constrained Tasks

• KA technique in which the expert performs a task
  they would normally do, but with constraints.
• Variations
• limited time
• limited data
• Useful for focusing the expert on essential
  knowledge and priorities

31
20-Questions

• KA technique in which the expert asks yes/no
  questions to the knowledge engineer in order to
  deduce an answer.
• The knowledge engineer need not know much about
  the domain, or have an answer in mind, just
  answer yes or no randomly
• The questions asked provide a good way of quickly
  acquiring attributes in a prioritised order.

32
Commentating and protocol generation

• KA technique in which the expert provides a
  running commentary of their own or anothers task
  performance.
• A valuable method for acquiring process knowledge
  and tacit knowledge.
• Variations
• self-reporting
• imaginary self-reporting
• self-retrospective
• shadowing
• retrospective shadowing

33
Teach back

• KA technique in which the knowledge engineer
  explains knowledge from part of the domain back
  to the expert.
• The expert then makes comments.
• Helps reveal misunderstandings and clarifies
  terminology.

34
Laddering

• KA technique that involves the construction,
  modification and validation of trees.
• A valuable method for acquiring concept knowledge
  and, to a lesser extent, process knowledge.
• Can make use of various trees
• concept tree
• composition tree
• attribute tree
• process tree
• decision tree
• cause tree

35
Laddering -Demonstration usingPC-PACK Laddering
Tool
36
KA Techniques3

• Modelling Techniques

37
Modelling Techniques

• KA techniques that use knowledge models as the
  focus for discussion, validation and modification
  of knowledge.
• Can use any form of model, but important types
  include
• process mapping
• concept mapping
• state diagram mapping

38
Process Mapping

• KA technique that involves the construction,
  modification and validation of process maps.
• A valuable method for acquiring process knowledge
  and tacit knowledge.

39
Process Map - Example
40
Process Mapping -Demonstration usingPC-PACK
Diagram Tool
41
Concept Mapping

• KA technique that involves the construction,
  modification and validation of concept maps.
• A good method for acquiring concept knowledge.

42
Concept Map - Example
written by
Author
is a
Oliver Twist
Charles Dickens
wrote
is a
wrote
admired
shorter than
is a
Dostoevsky
wrote on
Bleak House
is a
born in
Book
Russia
has part
Page
Paper
made from
43
State Diagram Mapping

• KA technique that involves the construction,
  modification and validation of a state diagram.
• A different approach to process mapping.
• Useful for capturing process knowledge, concept
  knowledge and tacit knowledge.

44
State Diagram - Example
Your number is dialed
On hook - no ringing
On hook - ringing
Lift receiver
Person at other end rings off
Lift receiver
Off hook - conversation
Off hook - dialing tone
Hang up
Phone is answered at other end
Hang up
Off hook - ringing tone
Dial number
Off hook - dialing
Complete dialing
45
How do you design a KA session plan?
Your 10 step guide
46
Designing a KA plan

• We need different techniques because
• there are different types of knowledge
• acquiring a certain type knowledge is made more
  efficient using the right technique
• e.g. can't get tacit knowledge using interviews
• Three types of KA techniques
• Natural (e.g. interviews, observation)
• Contrived (e.g. commentary, rep grid,
  20-questions)
• Modelling (e.g. process mapping)

47
Designing a KA Session Plan

• Be clear what knowledge you want from the
  session.
• Write an introduction summarising what knowledge
  you want from the session.
• Select the best KA technique/s to use.
• How do we do this? ..

48
Designing a KA Session Plan

• Place the techniques selected in a clear and
  logical order
• e.g. interview questions first
• e.g. commentary and protocols before process
  mapping
• Always end the session plan with the following
  question
• "Bearing in mind the goals of this session, what
  vital knowledge have we not yet covered"
• Assign timings to each section.

49
Designing a KA Session Plan

• If possible, check the session plan with your
  project manager or colleague and make amendments
  if necessary.
• Send (email, fax) the session plan to the expert
  at least one day before the session.
• Make any changes the expert suggests.
• During the session, stick to the plan and keep to
  the timings

50
Which KA technique to use

• Decide what type/s of conceptualisation and
  knowledge you need from the expert
• Is it structural objects oriented knowledge?
  (i.e. of concepts, attributes, states
  relationships)
• Is it process knowledge? (i.e. how to do things)
• Is it explicit knowledge? (i.e. easily explained)
• Is it tacit knowledge? (i.e. not easily
  explained)
• Use the diagram shown next to select the best
  technique/s to use..

51
Process Mapping
Observation
Process Knowledge
Protocols and Commentaries
State Diagram Mapping Teach Back
Constrained Tasks 20-questions
Repertory Grid
Interviews
Laddering
Triadic Method
Concept Knowledge
Concept Mapping
Card Sorting
Tacit Knowledge
Explicit Knowledge
52
Building an ontology a quick tutorial example

• Look at the following materials and consider how
  you might extract and model a conceptual model of
  part of this domain igneous rocks
• Structured interview
• Self report
• Repertory Grid
• Laddered Interview
• Item/Card Sort
• Photos thin and hand specimen

53
Ontology Capture

• Scope
• Brainstorm
• Group
• Main Phases Knowledge Acquisition
• Produce Definitions
• Do not commit to meta-ontology early
• Terms (proceed middle out)
• Concensus
• Handling ambiguity
• Guidelines
• Wording
• Review
• Meta-ontology

54
Ontology Documentation Skuces 4 Layer Model

• Level 0 Meta Assumptions
• MA 1 The ontology is divided into units termed
  categories that are hierarchically organised
• Level 1 Category Assumptions (4-tuple)
• Conceptual assumptions Explain what are the
  assumptions or rationale underlying the category.
  Why have such a category?
• Terminological Assumptions List the term or
  terms used. Explain why chosen. What terms in
  other language are equivalent
• Definitional Assumptions Define as in Dictionary
• Examples
• Level 2 NS Properties and Dimensions
• Level 3 Adding Non-logical Properties
• Typical, optional

55
D3E Discussion Spaces
56
Conceptualisation Model Pitfalls

• Pitfall Missing ontological elements
• Missing classes
• Missing attributes
• Confuse 11 with 1Many, or 1Many with ManyMany
• Important data is stored within text/comment
  fields
• Pitfall Extra ontological elements
• Pitfall Stop over-elaborating when do I stop?
• Proteins ? amino acid residues ? side chains ?
  physical chemical properties .
• Pitfall Relevance do I really need all this
  detail?
• Do we need to mention all the types of nucleic
  acid?

57
Integrating Existing Ontologies

• Reuse or adapt existing ontologies when possible
• Save time
• Correctness
• Facilitate interoperation
• Integration of ontologies
• Ontologies have to be aligned
• Hindered by poor documentation and argumentation
• Hindered by implicit assumptions
• Shared generic upper level ontologies should make
  integration easier

58
Encoding Implementation Toolkit

• Construct ontology using an ontology-development
  system
• Does the data model have the right expressivity?
• Is it just a taxonomy or are relationships
  needed?
• Is multiple parentage needed? Inverse
  relationships?
• What types of constraints are needed?
• Are reasoning services needed?
• What are authoring features of the development
  tool?
• Can ontology be exported to a DBMS schema?
• Can ontology be exported to an ontology exchange
  language?
• Is simultaneous updating by multiple authors
  needed?
• Size limitations of development tool?

59
Encoding Ontology Implementation Pitfalls

• Pitfall Semantic ambiguity
• Multiple ways to encode the same information
• Meaning of class definitions unclear
• Pitfall Encoding Bias
• Encoding the ontology changes the ontology
• Pitfall Redundancy (lack of normalization)
• Exact same information repeated
• Presence of computationally derivable information
• Date of birth and age
• DNA sequence and reverse complement
• More effort required for entry and update
• Partial updates lead to inconsistency
• OK if redundant information is maintained
  automatically

60
Encoding The Interaction Problem

• Task influences what knowledge is represented and
  how its represented
• Molecular biology chemical and physical
  properties of proteins
• Bioinformatics accession number, function gene
• Underlying perspectives mean they may not be
  reconcilable
• If an ontology has too many conflicting tasks it
  can end up compromised TAMBIS Ontology
  experience

61
Evaluate it - A guide for reusability

• Conciseness
• No redundancy Appropriateness protein
  molecules at the atomic resolution when amino
  acid level enough
• Clarity Consistency
• Satisfiability it doesnt contradict itself
• Enzyme is a both a protein which catalyses a
  reaction and does not catalyse a reaction
• Extensibility
• Minimal Commitment
• Do I have to buy into a load of stuff I dont
  really need or want just to get the bit I do?
• Minimal Encoding Bias

62
Documentation Make Ontology Understandable!

• Produce clear informal and formal documentation
• An ontology that cannot be understood will not be
  reused
• There exists a space of alternative ontology
  design decisions
• Semantics / Granularity
• Terminology
• Pitfall Neglecting to record design rationale

63
Publish the Ontology

• Formal and informal specifications
• Intended domain of application
• Design rationale
• Limitations

64
Further Reading

• ltto be suppliedgt