Transcript Guided Mediation for Agents (DOGMA) framework

EUROLAN 2003: VUB STAR Lab practical session
Peter Spyns & Jan De Bo
Vrije Universiteit Brussel – STAR Lab
Pleinlaan 2, Gebouw G-10
B-1050 Brussel, Belgium
http://www.starlab.vub.ac.be
Short introduction & overview
1. Developping Ontology Guided Mediation for Agents
(DOGMA) framework
Model theoretic (DB inspired)
Double articulation
Elementary (binary) fact types
Mining from texts, DB schemas, experts
Commitments (semantic constraints)
ORM style constraint definition
2. Modelling Exercice
Example
Manual
DogmaModeller tool
Lexons (1)
« … the means of transport that are currently available for a travel
agency. … the following ones: planes, trains, cars, ferries,
motorbikes and ships. »
travel_agency arranges/is_arranged_by trip
trip is_done_with/is_needed_for means_of_transport
plane is_kind_of/subsumes means_of_transport
train is_kind_of/subsumes means_of_transport
alternatively: set of values for means_of_transport
« …usually interested in the kind of planes …: is it a Boeing, or is it
an Airbus ? .. in the specific model … (a Boeing 717 or 777) »
company is_identified_by/identifies name
plane is_named_after/gives_its_name_to company
plane_model is_kind_of/comprises plane
plane_model is_identified_by/identifies type_number
Lexons (2)
« ... the travel agency is especially interested in flights, as it is the
means of transport mostly used by its customers »
travel_agency serves/books_with customer
customer travels_by/transports means_of_transport
travel_agency organises/is_organised_by flights
Commitments (1)
logic relations quantifiers – e.g., not, and, or, disjoint, each ,some
« each model of transport belongs only to one kind of
transportation (it’s either a plane, or a bus, etc.). »
 exclusive subtypes
instances and values – e.g., Italy, 1999
« Boeing 717 »
cause, reason and purpose – e.g., due to,
addition, exception and restriction – e.g., besides
condition, contrast and comparison – e.g., if…then
Modifiers:
adverbs – e.g., partly , wholly
« no other kinds of transport. »  disjoint mandatoriness
cardinality, frequency , modality
time relation, frequency – e.g., from…to…, weekly
Verbalisation
Semantic Constraints
• Uniqueness
internal
external
• Mandatoriness
Disjunctive or not
•
•
•
•
•
•
Subtype
Value
Subset
Exclusion
Equality
Frequency
Uniqueness
1:1  
one person is_husband_of/is_wife_of at most one person
1:m

each person is_husband_of/is_wife_of many person
each person is_wife_of/is_husband_of at most one person
m:1 
each person is_husband_of/is_wife_of at most one person
m:n --------
many person is_husband_of/is_wife_of many person
Mandatoriness
Each person has at least one person_name
Each person has exactly one person_name
(at least one + at most one = exactly one)

Each person studies and/or teaches a course (inclusive or)
Subset and Equality
If a person drives a car, then he owns a car
A person drives a car if and only if he drives a car
Exclusion
Each person studies or teaches a course (exclusive or)
Subtype
a person has to be male or female