From Tetherless World Wiki

To view this tutorial, you need to use Firfox and install Greasemonkey, and Wikipedia Presentation script. Don't be panic, those can be easily done by several mouse clicks. When you see "Start presentation" under the title, you are all set!

Introduction to Description Logics

Outline: Why we need DL and what is DL?

• Why we need DL and what is DL?
• The basic DL: ALC
• Extensions of ALC
• Tools and Resources
• DL Research at ISU

Logics and Ontology

_{(from [Wikipedia])}

• Logic : "is the study of the principles and criteria of valid inference and demonstration."
  • e.g., First Order Logics, Modal Logics
  • DL is a family of Logics

• Ontology: "is a data model that represents a set of concepts within a domain and the relationships between those concepts. It is used to reason about the objects within that domain."
  • e.g. Yellow Page Classification, Library Records, Contact Lists
  • DL is the foundation of current web ontology language (a.k.a. OWL).

Ontology: Examples

Ontology used in this Wiki contains:

• Concept inclusion: Research Associates are Person
• Concept instantiation: Jie Bao is an instance of Research Associate
• Concept Intersection: Jie Bao is Research Associate and Chinese
• Binary Relation: "Jie Bao" has affiliation "Artificial Intelligence Research Laboratory (ISU)"
  • See more
• Attributes of Instances: "Semantic Web Seminar (2007 Fall)" has duration "4 months" (time)

Ontology and Semantic Web

Task: find papers that have at least one author who is a Graduate student.

• Syntax search (with no ontology) :
  • "Paper Graduate Student" [1]
• Semantic search (with ontology):
  • [[Category:Paper]] [[has author::<q>[[Category:Graduate Student]]</q>]] [2]

• Semantics helps!

More Semantics in Need

Not Supported yet on this wiki!

• Concept Negation: a Man is not a Woman
• Concept Union: a Child is a Boy or a Girl
• Domain and Range restriction: if A has author B, then A is a Category:Document and B is a Category:Person
• Transitive Relation: if
  • Artificial Intelligence Research Laboratory (ISU) is a part of Department of Computer Science (ISU)
  • Department of Computer Science (ISU) is a part of Iowa State University
  • Then Artificial Intelligence Research Laboratory (ISU) is a part of Iowa State University
• Inverse Relation:
  • If Jie Bao participates in INDUS(Project)
  • Then INDUS(Project) has participant Jie Bao
• Cardinality Restriction: Every Paper has at least one author

Description Logics

• Allows expressive semantic tags on the web (as shown in the last slide)
• Formal semantics
• Has decidability (can always answer a query in finite steps)
• Has available tools (e.g., editors, parsers, and reasoners)
• Is the foundation of web ontology languages, e.g., OWL

Outline: The basic DL: ALC

• Why we need DL and what is DL?
• The basic DL: ALC
• Extensions of ALC
• Tools and Resources
• DL Research at ISU

Basic Elements in DL

• Concept (class): e.g., Person
• Role (property): e.g., has author, has title
• Instance (individual): e.g. Jie Bao
• Everything (top): Failed to parse (Missing texvc executable;

please see math/README to configure.): \top

• Nothing (bottom): Failed to parse (Missing texvc executable;

please see math/README to configure.): \bot

ALC

TBox and ABox

An ALC ontology (i.e., knowledge base) may has a Tbox and an ABox

• TBox (terminology box): the "schema" of the ontology, e.g., Concept hierarchy of this wiki
  • General Concept Inclusion (GCI), e.g., Failed to parse (Missing texvc executable;

please see math/README to configure.): \mathsf{Student}\sqsubseteq\mathsf{Person}

• ABox (assertion box): the instance set of the ontology, e.g., All objects in this wiki
  • Failed to parse (Missing texvc executable;

please see math/README to configure.): \mathsf{Person}(\mathsf{Jie Bao})

or Failed to parse (Missing texvc executable;

please see math/README to configure.): \mathsf{Jie Bao}:\mathsf{Person}

• • Failed to parse (Missing texvc executable;

please see math/README to configure.): \mathsf{hasPet}(\mathsf{Jie Bao},\mathsf{George})

or Failed to parse (Missing texvc executable;

please see math/README to configure.): (\mathsf{Jie Bao},\mathsf{George}):\mathsf{hasPet}

     -- George?

Reasoning

• Concept Subsumption: e.g., if Failed to parse (Missing texvc executable;

please see math/README to configure.): \exists \mathsf{hasChild}.\mathsf{Man}\sqsubseteq \exists \mathsf{hasChild}.\mathsf{Human}

always holds?

• Concept Satisfability: e.g., if Failed to parse (Missing texvc executable;

please see math/README to configure.): \exists \mathsf{hasChild}.\mathsf{Man}\sqcap \forall \mathsf{hasChild}.\mathsf{Woman}

may have an instance?

• Note that: Failed to parse (Missing texvc executable;

please see math/README to configure.): C\sqsubseteq D \Leftrightarrow unsat(C\sqcap \neg D)

• Modern DLs use Tableau Algorithm for reasoning.

Outline: Extensions of ALC

• Why we need DL and what is DL?
• The basic DL: ALC
• Extensions of ALC
• Tools and Resources
• DL Research at ISU

Extensions of ALC

and Failed to parse (Missing texvc executable;

with transitive properties.

DL vs. Other Ontology Languages

Outline: Tools and Resources

• Why we need DL and what is DL?
• The basic DL: ALC
• Extensions of ALC
• Tools and Resources
• DL Research at ISU

Editors

• Protege (open source)
• Swoop (open source)
• Topbraid Composer

All Java-based

Reasoner

• CEL: LISP-based, for simple ontologies
• FaCT++: C++-based
• KAON2 Java-based, implemented using resolution methods.
• Pellet Java-based
• RacerPro lisp-based

Key References

• Key papers to read
• Tutorials
  • Uli Sattler's video tutorial
  • DL Workshop and other tutorials

Outline: DL Research at ISU

• Why we need DL and what is DL?
• The basic DL: ALC
• Extensions of ALC
• Tools and Resources
• DL Research at ISU

DL Research at ISU

• People: Vasant Honavar, Giora Slutzki, Jie Bao, Doina Caragea, George Voutsadakis
• Research Topics: (Papers)
  • Modular Ontologies using Package-based Description Logics
  • Distributed Reasoning with ontologies
  • Privacy-Preserving Reasoning on the Semantic Web
• Implementation in action
  • Modular ontology editor: COB-Editor
  • Distributed Reasoner
  • This wiki (a universal testbed)

Funding

• Funding: NSF IIS-0639230 SGER: Exploratory Investigation of Modular Ontologies. National Science Foundation Vasant Honavar (PI), Giora Slutzki (Co-PI), and Doina Caragea (Co-PI).
• A Collaborative Building Environment for Animal Trait Ontologies, Vasant Honavar (PI), James Reecy, (Co-PI), Center for Integrative Animal Genomics and Center for Computational Intelligence, Learning, and Discovery, Iowa State University.

Facts About the Presentation