kbs
play

KBS Knowledge-Based Systems Group The need of common-sense - PowerPoint PPT Presentation

Realizing Default Logic over Description Logic Knowledge Bases Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner KBS Group, Institute of Information Systems, Vienna University of Technology ECSQARU 2009 July 2, 2009 KBS Knowledge-Based


  1. Realizing Default Logic over Description Logic Knowledge Bases Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner KBS Group, Institute of Information Systems, Vienna University of Technology ECSQARU 2009 — July 2, 2009 KBS Knowledge-Based Systems Group

  2. The need of common-sense reasoning on top of ontologies Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 2/17

  3. The need of common-sense reasoning on top of ontologies Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 2/17

  4. The need of common-sense reasoning on top of ontologies Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 2/17

  5. The need of common-sense reasoning on top of ontologies default reasoning on top of ontologies? Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 2/17

  6. The need of common-sense reasoning on top of ontologies default reasoning on top of ontologies? integrations of rules and ontologies : cq-programs Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 2/17

  7. Description Logic Knowledge Bases (DL-KBs) Syntax and Semantics Name Syntax Semantics ∆ I / ∅ ⊤ / ⊥ Top/Bottom C I ∩ D I C ⊓ D Intersection C I ∪ D I C ⊔ D Union ∆ I \ C I ¬ C Negation { a ∈ ∆ I | ∀ b . ( a , b ) ∈ R I → b ∈ C I } ∀ R . C Value restriction { a ∈ ∆ I | ∃ b . ( a , b ) ∈ R I ∧ b ∈ C I } ∃ R . C Existential quant. Modeling : TBox & ABox Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 3/17

  8. Description Logic Knowledge Bases (DL-KBs) Syntax and Semantics Name Syntax Semantics ∆ I / ∅ ⊤ / ⊥ Top/Bottom C I ∩ D I C ⊓ D Intersection C I ∪ D I C ⊔ D Union ∆ I \ C I ¬ C Negation { a ∈ ∆ I | ∀ b . ( a , b ) ∈ R I → b ∈ C I } ∀ R . C Value restriction { a ∈ ∆ I | ∃ b . ( a , b ) ∈ R I ∧ b ∈ C I } ∃ R . C Existential quant. Modeling : TBox & ABox Translation to first-order logic π x ( A ) = A ( x ) π x ( C ⊓ D ) = π x ( C ) ∧ π x ( D ) π x ( ∀ R . C ) = ∀ y . R ( x , y ) ⊃ π y ( C ) π x ( C ⊔ D ) = π x ( C ) ∨ π x ( D ) π x ( ∃ R . C ) = ∀ y . R ( x , y ) ∧ π y ( C ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 3/17

  9. Default Theories over DL-KBs ∆ = � L , D � similar to [Baader and Hollunder, 1995] Default rule α ( � β i ( � X ) Y i ) � �� � � �� � α 1 ( � X 1 ) ∧ · · · ∧ α k ( � X k ) : β 1 ( � β i , 1 ( � Y i , 1 ) ∧ · · · ∧ β i ,ℓ i ( � Y i ,ℓ i ) , . . . , β m ( � Y 1 ) , . . . , Y m ) γ 1 ( � Z 1 ) ∧ · · · ∧ γ n ( � Z n ) � �� � γ ( � Z ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 4/17

  10. Default Theories over DL-KBs ∆ = � L , D � similar to [Baader and Hollunder, 1995] Default rule α ( � β i ( � X ) Y i ) � �� � � �� � α 1 ( � X 1 ) ∧ · · · ∧ α k ( � X k ) : β 1 ( � β i , 1 ( � Y i , 1 ) ∧ · · · ∧ β i ,ℓ i ( � Y i ,ℓ i ) , . . . , β m ( � Y 1 ) , . . . , Y m ) γ 1 ( � Z 1 ) ∧ · · · ∧ γ n ( � Z n ) � �� � γ ( � Z ) Semantics: based on the Γ ∆ operator ◮ Let S be a set of assertions, then Γ ∆ ( S ) is the smallest set that ◮ contains Cn ( L ) ◮ is deductively closed ◮ if α ( � X ) ∈ Γ ∆ ( S ) and ¬ β i ( � ∈ S , then γ ( � Y i ) / Z ) ∈ Γ ∆ ( S ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 4/17

  11. Default Theories over DL-KBs ∆ = � L , D � similar to [Baader and Hollunder, 1995] Default rule α ( � β i ( � X ) Y i ) � �� � � �� � α 1 ( � X 1 ) ∧ · · · ∧ α k ( � X k ) : β 1 ( � β i , 1 ( � Y i , 1 ) ∧ · · · ∧ β i ,ℓ i ( � Y i ,ℓ i ) , . . . , β m ( � Y 1 ) , . . . , Y m ) γ 1 ( � Z 1 ) ∧ · · · ∧ γ n ( � Z n ) � �� � γ ( � Z ) Semantics: based on the Γ ∆ operator ◮ Let S be a set of assertions, then Γ ∆ ( S ) is the smallest set that ◮ contains Cn ( L ) ◮ is deductively closed ◮ if α ( � X ) ∈ Γ ∆ ( S ) and ¬ β i ( � ∈ S , then γ ( � Y i ) / Z ) ∈ Γ ∆ ( S ) ◮ E is an extension of ∆ iff Γ ∆ ( E ) = E Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 4/17

  12. Example ∆ = � L , D � � Flier ⊑ ¬ NonFlier , Penguin ⊑ Bird , � L = Penguin ⊑ NonFlier , Bird ( tweety ) � Bird ( X ) : Flier ( X ) � D = Flier ( X ) E = Cn ( L ∪ { Flier ( tweety ) } ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 5/17

  13. Example ∆ ′ = � L ′ , D � � Flier ⊑ ¬ NonFlier , Penguin ⊑ Bird , � L ′ = Penguin ⊑ NonFlier , Penguin ( tweety ) � Bird ( X ) : Flier ( X ) � D = Flier ( X ) E ′ = Cn ( L ′ ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 6/17

  14. Conjunctive Query Programs [Eiter et al. , 2008a] ◮ (union of) conjunctive queries: q ( X ) = { X | AirCraft ( X ) ∨ ( UFO ( X ) ∧ ¬ Hoax ( X )) } Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 7/17

  15. Conjunctive Query Programs [Eiter et al. , 2008a] ◮ (union of) conjunctive queries: q ( X ) = { X | AirCraft ( X ) ∨ ( UFO ( X ) ∧ ¬ Hoax ( X )) } ◮ cq-atom: DL [ AirCraft ⊎ isBoeing ; AirCraft ( X ) ∨ ( UFO ( X ) , ¬ Hoax ( X ))]( X ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 7/17

  16. Conjunctive Query Programs [Eiter et al. , 2008a] ◮ (union of) conjunctive queries: q ( X ) = { X | AirCraft ( X ) ∨ ( UFO ( X ) ∧ ¬ Hoax ( X )) } ◮ cq-atom: DL [ AirCraft ⊎ isBoeing ; AirCraft ( X ) ∨ ( UFO ( X ) , ¬ Hoax ( X ))]( X ) ◮ cq-rule: flying thing ( X ) ← thing ( X ) , DL [ AirCraft ⊎ isBoeing ; AirCraft ( X ) ∨ ( UFO ( X ) , ¬ Hoax ( X ))]( X ) . Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 7/17

  17. Conjunctive Query Programs [Eiter et al. , 2008a] ◮ (union of) conjunctive queries: q ( X ) = { X | AirCraft ( X ) ∨ ( UFO ( X ) ∧ ¬ Hoax ( X )) } ◮ cq-atom: DL [ AirCraft ⊎ isBoeing ; AirCraft ( X ) ∨ ( UFO ( X ) , ¬ Hoax ( X ))]( X ) ◮ cq-rule: flying thing ( X ) ← thing ( X ) , DL [ AirCraft ⊎ isBoeing ; AirCraft ( X ) ∨ ( UFO ( X ) , ¬ Hoax ( X ))]( X ) . ◮ cq-program: KB = ( L , P ) — based on answer set semantics Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 7/17

  18. Answer Set Semantics of cq-Programs generalized [Gelfond and Lifschitz, 1991] P is a set of rules of form a ← b 1 , . . . , b m , not c 1 , . . . , not c n . Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 8/17

  19. Answer Set Semantics of cq-Programs generalized [Gelfond and Lifschitz, 1991] P is a set of rules of form a ← b 1 , . . . , b m , not c 1 , . . . , not c n . ◮ Interpretation I ⊆ HB P Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 8/17

  20. Answer Set Semantics of cq-Programs generalized [Gelfond and Lifschitz, 1991] P is a set of rules of form a ← b 1 , . . . , b m , not c 1 , . . . , not c n . ◮ Interpretation I ⊆ HB P ◮ I is an answer set of P if I is the least model of the reduct P I Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 8/17

  21. Answer Set Semantics of cq-Programs generalized [Gelfond and Lifschitz, 1991] P is a set of rules of form a ← b 1 , . . . , b m , not c 1 , . . . , not c n . ◮ Interpretation I ⊆ HB P ◮ I is an answer set of P if I is the least model of the reduct P I P I is constructed by ◮ removing rules r ∈ P such that c i ∈ I ◮ removing all c i and nonmonotonic cq-atoms from remaining rules Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 8/17

  22. Transformation Ω (inspired by [Eiter et al. , 2008b]) D is a set of defaults of form β i ( � Y i ) � �� � α ( � X ) : β 1 ( � β i , 1 ( � Y i , 1 ) ∧ · · · ∧ β i ,ℓ i ( � Y i ,ℓ i ) , . . . , β m ( � Y 1 ) , . . . , Y m ) γ 1 ( � Z 1 ) ∧ · · · ∧ γ n ( � Z n ) � �� � γ ( � Z ) Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 9/17

  23. Transformation Ω (inspired by [Eiter et al. , 2008b]) D is a set of defaults of form β i ( � Y i ) � �� � α ( � X ) : β 1 ( � β i , 1 ( � Y i , 1 ) ∧ · · · ∧ β i ,ℓ i ( � Y i ,ℓ i ) , . . . , β m ( � Y 1 ) , . . . , Y m ) γ 1 ( � Z 1 ) ∧ · · · ∧ γ n ( � Z n ) � �� � γ ( � Z ) R = { in γ i ( � Z i ) ← in γ ( � ◮ Concluding rules Z ) | 1 ≤ i ≤ n } Minh Dao-Tran, Thomas Eiter, Thomas Krennwallner Realizing Default Logic over Description Logic 9/17

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend