1 Letting the classifier check your intuitions Existentials, Universals, & other logical variants Some, Only, Not, And, Or, etc. Lab exercise - 3b Alan.

Slides:



Advertisements
Similar presentations
Charting the Potential of Description Logic for the Generation of Referring Expression SELLC, Guangzhou, Dec Yuan Ren, Kees van Deemter and Jeff.
Advertisements

29/11/2004CS646: N. Drummond, M. Horridge1 Presented by the CO-ODE and HyOntUse projects Funded by CS646: OWL Tutorial (session 2)
Artificial Intelligence Description Logics - Background and Motivation for DL Relation between DL and FOPL (extensional Semantics) Example.
ARCHITECTURES FOR ARTIFICIAL INTELLIGENCE SYSTEMS
OWL - DL. DL System A knowledge base (KB) comprises two components, the TBox and the ABox The TBox introduces the terminology, i.e., the vocabulary of.
Algebra Problems… Solutions Algebra Problems… Solutions © 2007 Herbert I. Gross Set 13 By Herbert I. Gross and Richard A. Medeiros next.
“Elephant Traps” Alan Rector with The Bio Health Informatics Group The CO-ODE Team University of Manchester With acknowledgements to Natsha Noy and the.
Knowledge Representation
Based on “A Practical Introduction to Ontologies & OWL” © 2005, The University of Manchester A Practical Introduction to Ontologies & OWL Session 3: Additional.
1 OWL Pizzas: Practical Experience of Teaching OWL-DL: Common Errors & Common Patterns Alan Rector 1, Nick Drummond 1, Matthew Horridge 1, Jeremy Rogers.
Tutorial Le Phuoc Son Hoang Huu Hanh Hue University.
Of 27 lecture 7: owl - introduction. of 27 ece 627, winter ‘132 OWL a glimpse OWL – Web Ontology Language describes classes, properties and relations.
Ontologies - Design principles Cartic Ramakrishnan LSDIS Lab University of Georgia.
Copyright © Cengage Learning. All rights reserved. CHAPTER 2 THE LOGIC OF COMPOUND STATEMENTS THE LOGIC OF COMPOUND STATEMENTS.
The Semantic Web Week 15 Reasoning with (and Visualising) Ontologies Module Website: Practical :Protégé-2000 WITH.
USCISIUSCISI Loom: Basic Concepts Thomas A. Russ USC Information Sciences Institute.
Part 6: Description Logics. Languages for Ontologies In early days of Artificial Intelligence, ontologies were represented resorting to non-logic-based.
THE TRANSITION FROM ARITHMETIC TO ALGEBRA: WHAT WE KNOW AND WHAT WE DO NOT KNOW (Some ways of asking questions about this transition)‏
Algebra Problems… Solutions
Copyright © Cengage Learning. All rights reserved.
DAML+OIL Ontology Tutorial Chris Wroe, Robert Stevens (Sean Bechhofer, Carole Goble, Alan Rector, Ian Horrocks….) University of Manchester.
1 Foundations of the Semantic Web: Ontology Engineering Building Ontologies 1 Alan Rector & colleagues.
4-1 Identifying Linear Functions
Protege OWL Plugin Short Tutorial. OWL Usage The world wide web is a natural application area of ontologies, because ontologies could be used to describe.
1 Making OWL Easier: Practical Ontology Development in using Protégé-OWL-CO-ODE Tools Alan Rector, Hai Wang, Jeremy Rogers with acknowledgement to Nick.
Math 3121 Abstract Algebra I Section 0: Sets. The axiomatic approach to Mathematics The notion of definition - from the text: "It is impossible to define.
Section 3.3a!!!. First, remind me… What does the horizontal line test tell us??? More specifically, what does it tell us about the function This function.
Building an Ontology of Semantic Web Techniques Utilizing RDF Schema and OWL 2.0 in Protégé 4.0 Presented by: Naveed Javed Nimat Umar Syed.
USCISIUSCISI Background Description Logic Systems Thomas Russ.
Based on “A Practical Introduction to Ontologies & OWL” © 2005, The University of Manchester A Practical Introduction to Ontologies & OWL Session 1: Primitive.
OWL 2 in use. OWL 2 OWL 2 is a knowledge representation language, designed to formulate, exchange and reason with knowledge about a domain of interest.
1 Inference Rules and Proofs (Z); Program Specification and Verification Inference Rules and Proofs (Z); Program Specification and Verification.
BioHealth Informatics Group Ontology Tutorial, © 2005 Univ. of Manchester1 Informal Modelling Robert Stevens.
BioHealth Informatics Group A Practical Introduction to Ontologies & OWL Session 2: Defined Classes and Additional Modelling Constructs in OWL Nick Drummond.
Coastal Atlas Interoperability - Ontologies (Advanced topics that we did not get to in detail) Luis Bermudez Stephanie Watson Marine Metadata Interoperability.
Computing & Information Sciences Kansas State University Wednesday, 20 Sep 2006CIS 490 / 730: Artificial Intelligence Lecture 12 of 42 Wednesday, 20 September.
Logic …. Disjoint Properties As for disjoint classes, two properties can be disjoint (owl : propertyDisjointWith) Property p and p’ are disjoint if no.
Based on “A Practical Introduction to Ontologies & OWL” © 2005, The University of Manchester A Practical Introduction to Ontologies & OWL Session 2: Defined.
© O. Corcho, MC Suárez de Figueroa Baonza 1 OWL and SWRL Protégé 4: Building an OWL Ontology Mari Carmen Suárez-Figueroa, Oscar Corcho {mcsuarez,
OilEd An Introduction to OilEd Sean Bechhofer. Topics we will discuss Basic OilEd use –Defining Classes, Properties and Individuals in an Ontology –This.
2nd Feb 2005Protege-OWL tutorial, © 2005 Univ. of Manchester1 Protégé-OWL Tutorial Session 2: Defined Classes Nick Drummond.
2nd Sept 2004UK e-Science all hands meeting1 Designing User Interfaces to Minimise Common Errors in Ontology Development Alan Rector, Nick Drummond, Matthew.
Week 2 Introduction to Data Modelling
Ontology Engineering Lab #5 – September 30, 2013.
Ontology Engineering Lab #3 – September 16, 2013.
1 Georgia Tech, IIC, GVU, 2006 MAGIC Lab Rossignac Lecture 02: QUANTIFIERS Sections 1.3 and 1.4 Jarek Rossignac CS1050:
ONTOLOGY ENGINEERING Lab #3 – September 15,
Description Logics Dr. Alexandra I. Cristea. Description Logics Description Logics allow formal concept definitions that can be reasoned about to be expressed.
ece 627 intelligent web: ontology and beyond
Ontology Engineering Lab #4 - September 23, 2013.
1 Copyright, 1996 © Dale Carnegie & Associates, Inc. Chapter 2: Logic & Incidence Geometry Back To the Very Basic Fundamentals.
BioHealth Informatics Group 1 Techniques for segmenting large description logic ontologies Julian Seidenberg, Alan Rector
Enable Semantic Interoperability for Decision Support and Risk Management Presented by Dr. David Li Key Contributors: Dr. Ruixin Yang and Dr. John Qu.
BioHealth Informatics Group Copyright © 2005, The University of Manchester1 A Practical Introduction to Ontologies & OWL Additional Exercises: Common Errors.
Of 29 lecture 15: description logic - introduction.
LDK R Logics for Data and Knowledge Representation Description Logics: family of languages.
Ccs.  Ontologies are used to capture knowledge about some domain of interest. ◦ An ontology describes the concepts in the domain and also the relationships.
Chapter 1 Expressions, Equations and Inequalities.
“It is impossible to define every concept.” For example a “set” can not be defined. But Here are a list of things we shall simply assume about sets. A.
ece 720 intelligent web: ontology and beyond
Can SNOMED CT be harmonized with an upper-level ontology?
Lab exercise - 3a Alan Rector & colleagues
Chapter 2 Sets and Functions.
Session 2: Defined Classes Nick Drummond
CS646: OWL Tutorial (session 2)
ece 720 intelligent web: ontology and beyond
Session 1: Primitive Classes Nick Drummond
Foundations of the Semantic Web: Ontology Engineering
Description Logics.
Propositional Logic 1) Introduction Copyright 2008, Scott Gray.
Presentation transcript:

1 Letting the classifier check your intuitions Existentials, Universals, & other logical variants Some, Only, Not, And, Or, etc. Lab exercise - 3b Alan Rector & colleagues

2 To Submit (With Lab 4) A text report explaining the results –In six part as described here. Each clearly headed and with cross links to the ontology An ontology with branches demonstrating each of the issues –Or several ontologies clearly labelled to correspond to the report. All zipped with the results of the University Untangling exercise with Lab 4.

3 Closure Restrictions A standard pattern aProperty SOME Class 1 and aProperty SOME Class 2 and aProperty SOME Class 3 and … aProperty SOME Class n and aProperty ONLY (Class 1 or Class 2 or Class 3 or … or Class n ) disjoint (Class 1 Class 2 Class 3 … Class n closure restriction

4 Part 1: Create an new ontology with the following definitions Primitive Concepts –Family Child Girl Boy Properties –has_children Defined concepts –All_girl_family ≡ Family & (restriction has_children ONLY Girl) –No_child_family ≡ Family & not (restriction has_children SOME Top) –Zero_child_family ≡ Family & (restriction has_children max-cardinality=0 )

5 Classify it and examine the result What does this tell you about the use of ONLY (Universal qualification) Why do we say that most ontologies are based on SOME (Existential qualification) Repair the definition of “All_girl_family” so that it behaves as expected Make this “Result 1” in your report

6 Paraphrases Note: in what follows ‘THAT’ transforms to logical ‘AND’, i.e.: –Dog THAT has_owner SOME Person is the same as –DOG AND has_owner SOME Person

7 Part 2: Functional Properties with SOME & ONLY Create the classes Dog, Cat, and Person and make them disjoint Create the property ‘has_owner’ with the inverse ‘owns’ –Make has_owner functional -(tick functional box on properties tab) Define the four classes –Dog THAT has_owner SOME Person –Dog THAT has_owner ONLY Person –Person THAT owns SOME Dog –Person THAT owns ONLY Dog Run the classifier and explain the result. –Why is the pattern the subsumptions of Person and Dog different? What does this tell you about the use of SOME and ONLY with functional properties. –Why is this true? Submit this as “Result 2” in your report

8 Part 3: Conjunction, disjunction & negation (You should already have created the classes Dog, Cat, and Person and made them disjoint) Define - using necessary & sufficient conditions - the four classes –Person THAT owns SOME (Dog AND Cat) Person THAT owns ONLY (Dog AND Cat) Person THAT owns SOME (Dog OR Cat) Person THAT owns ONLY (Dog OR Cat) Before you run the classifier, predict which will be satisfiable and how the inferred subsumption hierarchy will be arranged. Run the classifier and check your understanding. Write up the explanation for the result Set a restriction on Person that every person owns at least one thing (owns min 1) –Predict how will this affect the classification. –Run the classifier. Explain the result. - it may be easier to see in OWLViz)

9 Negation Create examples with the negation of the previous restrictions - be sure to use necessary and sufficient conditions and leave the restriction that a person must own at least one thing. –Person THAT NOT owns SOME (Dog AND Cat) Person THAT NOT owns ONLY (Dog AND Cat) Person THAT NOT owns SOME (Dog OR Cat) Person THAT NOT owns ONLY (Dog OR Cat) –Before you run the classifier, predict which will be satisfiable and how the inferred subsumption hierarchy will be arranged. Run the classifier and check your understanding. Write up the explanation for the result. (some of these are hard) –Set the range of owns to (Dog OR Cat). How will this change the above results? Run the classifier, and explain. –Remove then restriction that a Person owns at least one thing. Explain the result –(It may be easier to see complex hierarchies in OWLViz). Also you might want to have a Person_1 and Person_2, owns_1, and owns_2 for different demonstrations.

10 Negation (cont) Add Person THAT (owns SOME Dog) AND (owns SOME Cat) Person THAT (owns SOME Dog) Person THAT (owns SOME Cat) Person THAT (owns min 2) Predict how these will classify with respect to the previous definitions. Classify and explain. Explain the relationship between: –owns SOME (Dog OR Cat) (owns SOME DOG) AND (owns SOME Cat) owns SOME (Dog AND Cat) Label the discussion for this section Results 3 in your report.

11 Part 4: Domain and Range Constraints If you have not already done so, create the class Animal and make it a superclass of both Cat and Dog. Create a sibling class Manufactured_thing so the hierarchy looks like –… Animal Manufactured_thing –Do NOT make Manufactured_thing and Animal disjoint yet. Make Car a primitive subclas of Manufactured_thing Make the domain of the property owns to be Person and the range to the Animal Define Personal_car as a car owned by a person and a Car_owner as a person that owns a car. Predict what will happen. Use the classifier to check Now make Animal and Manufactured_thing disjoint Predict what will happen and verify with the classifier Explain in write up under Results 4.

12 Part 5: General Inclusion Axioms Create defined class “Animals that have claws and meow” and create a necessary statement that it implies Cat. Add a restriction to Dog that all dogs have claws Create a subclass of dog that meows Predict the result; classify it as check. How would you debug this case? How might you make the errors easier to find? Submit as Results 5.

13 Part 6: Parts and Wholes Add to mammals that they all have four limbs Create a common parent for cat and dog - call it Cat_or_Dog Add that the limbs of Cat_or_Dog are of type leg. Add that all ‘digits’ (fingers and toes) are parts of legs of Cat_or_Dog Add that claws are parts of the digits of Cat_or_Dog. Create the ‘Adapted SEP Triples’ to show the parts hierarchy for Cat_of_Dog. Classify it and check. Explain why this is done with is_part_of rather than has_part. (You might want to demonstrate by experiment) Label as Results 6.