Stuart Aitken Artificial Intelligence Applications Institute
1 Artificial Intelligence Applications Institute Centre for Intelligent Systems and their...
-
Upload
joanna-miles -
Category
Documents
-
view
229 -
download
3
Transcript of 1 Artificial Intelligence Applications Institute Centre for Intelligent Systems and their...
1
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Stuart Aitken
Artificial Intelligence Applications Institute
A Process Ontology for A Process Ontology for Cell BiologyCell Biology
2
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
OutlineOutline
• Rapid Knowledge Formation (RKF) Project– RKF Project goals and domain– The Cyc knowledge based-system– RKF Tools
• Process Ontology– General approach– Formalisation– Example
3
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Rapid Knowledge FormationRapid Knowledge Formation
• The RKF project aims to develop tools which will allow domain experts to enter knowledge directly into the KBS.
• DARPA-funded, two teams:– CYCORP– SRI
• Organised around ‘Challenge Problems’ – Cell Biology
4
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
RKFRKF
Aim: To enable biologists to construct an ontology/KB from a textbook source
formalise
Ontology
Alberts et al, Essential Cell Biology, 1998
5
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Rapid Knowledge FormationRapid Knowledge Formation
Key techniques:• The KBS has knowledge of the KA
process– Knowledge of salience– Knowledge of the requirements of an
adequate formalisation
• There is a dialogue between expert and system, which clarifies the concept being defined.
6
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Rapid Knowledge FormationRapid Knowledge Formation
Evaluation:
After a period of tool development,• trials are organised, both• expert performance, and• KE performance is measured,• and assessed independently.
The evaluation is extensive – over a period of 2 weeks
7
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
The Cyc KBSThe Cyc KBS
• Cyc (Doug Lenat) is a knowledge-based system, under development since ~1984, aiming to represent common sense knowledge.
• Cyc uses a large upper-level ontology
• Uses a logical language based on first-order logic
8
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
The Cyc KBSThe Cyc KBS
Concepts in the Upper Ontology:– Thing, Agent, Event– TangibleThing, InformationBearingObject– …. Dog, Book– subclass(genls), instance-of(isa)– parts, subevent, role predicates– 1600 concepts in total in the public
release (1998) - small% of Cyc
Classification:– Stuff-like vs Object-like– Individual vs Set
9
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
The Cyc KBSThe Cyc KBS
• The upper-ontology supports application development:
Upper-level
Intermediate-level
Application-level
Thing
10
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
The Cyc KBSThe Cyc KBS
Cyc includes:• An inference engine, • GUI, • tools for ontology development.• Until the RKF project, ontology
development was by trained knowledge engineers, working with domain experts.
11
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
RKFRKF
New tools in Cyc:• Define a new concept, and place it
correctly in the ontology• Refine a concept definition• Define a new predicate• Assert a new fact• Define a new rule• State an analogy• Construct a new process
12
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
RKFRKF
User interaction:• Selection of items in the interface
– Choice determined ‘intelligently’, KBS has knowledge of salience, and the KA process, this knowledge must be authored
• Browsing of the ontology• Search• Natural language dialogue
13
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Process ModelsProcess Models
BindsTogether Move
RNA Transcription
14
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Process DescriptorProcess Descriptor
Q: Name the processA: [ RNA Transcription ]Q:Select the type of Process that describes
the category best• event localised• creation or destruction event…• ‘say this:’[ _ _ _ _ _ _ ]Q: Define:• affected object: [ _ _ _ _ _ ]• location: [ _ _ _ _ _ ]• actor: [ _ _ _ _ _ ]
15
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Process ModelsProcess Models
Describing Processes:• Complex expressions at the instance level• Simpler to describe in terms of types
Upper-level
Intermediate-level
subevent(Event,Event)doneBy(Event,Agent)
ForAll ?E ?F ?G implies(subevent(?E,?G) and isa(?E,BindsTogether)subevent(?F,?G) and isa(?F,Move))before(startOf(?E),startOf(?F))
Application-level?
16
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Script VocabularyScript Vocabulary
The Script theory defines the semantics of Type-Level assertions
(typePlaysRoleInScene RNATranscription DNAMolecule BindsTogether objectActedOn)
• Requires rules for identity– Can require complex reasoning
• Good for user input• Can be extended to cover pre and
postconditions of actions
17
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
ScriptsScripts
subevents
BindsTogether
e
Move
f
RNA Transcription
Forall subevents f of t, of type Move,and all subevents e of t, of type BindsTogether,(startsAfterStartingof f e) where t is of type RNATranscription
t
startsAfterStartingOfInScript
18
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
ScriptsScripts
Type playing role
N
BindsTogetherNucleotide
e
Types:
objectActedOn
Instance:
For some n in N, (objectActedOn e n)
19
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
New Script VocabularyNew Script Vocabulary
• Pre and Post conditions
BindsTogether
N
R
N
RnottouchingDirectly connectedTo
(preconditionOfScene-negated BindsTogether touchingDirectly <Ribonucleotide Nucleotide>)
(postconditionOfScene BindsTogether connectedTo <Ribonucleotide Nucleotide>)
20
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
New Script VocabularyNew Script Vocabulary
N R
Some ?n in N, some ?r in R(not(touchingDirectly ?n ?r))
Some ?n in N, some ?r in R(connectedTo ?n ?r)
BindsTogetherNucleotide Ribonucleotide
e
Types:
roleroleSet ofInstances:
Precondition: Postcondition:
identity
21
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Script VocabularyScript Vocabulary
• The Script vocabulary forms an ‘intermediate level’, which
• lies behind the Process descriptor GUI (i.e. the textboxes)
• Not, in itself, a taxonomy of processes, but allows processes to be described in detail.
• Defining the subclass relation is just one task.
22
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Vaccinia Virus Life CycleVaccinia Virus Life Cycle
• The vaccinia virus life cycle was selected as an example of a complex model to formalise as a set of Scripts.
• The model includes actions, decomposition, ordering, objects-playing-roles and pre/postconditions
• It is a good test for the Script vocabulary
23
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
Vaccinia Virus Life CycleVaccinia Virus Life Cycle
mRNATranscription-Early
ViralGeneTranslation-Early
MovementOfProtein
Temporal:
mRNATranscription-Early
ViralGeneTranslation-Early
MovementOfProtein
mRNATranscription-Early
ViralGeneTranslation-Early
MovementOfProtein
Participants
Conditions:
Outputs:messengerRNA
Inputs:messengerRNA
Pre:spatiallySubsumes Cell VirusCore
Post:spatiallySubsumes CellCytoplasm Vitf2
24
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
EvaluationEvaluation
• 8 biologists were selected, and trained in the tools, 4 per team
• The knowledge to be formalised was selected (chapter 7 in Alberts)
• The knowledge base was allowed to contain ‘pump-priming’ knowledge
• The biologists entered knowledge , using the tools, then tested it against a set of questions,
• Ontology/KB was revised
25
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
EvaluationEvaluation
Results (outline)• A huge amount of data was collected,
but analysis is complex (IET Inc)• Domain experts were able to develop
ontologies after ‘light’ training• Knowledge engineers out-perform
domain experts in ontology construction
26
Artificial Intelligence Applications InstituteCentre for Intelligent Systems and their Applications
SummarySummary
‘Power Tools’ for ontology development are being implemented and tested in the RKF project.
• A Script/Process vocabulary has been developed and applied to processes in cell biology, covering:– Temporal order– Participants– Pre/postconditions– Repetition