41
Updating TBoxes in DL-Lite D. Zheleznyakov

Updating TBoxes in DL-Lite

  • Upload
    kedem

  • View
    29

  • Download
    0

Embed Size (px)

DESCRIPTION

Updating TBoxes in DL-Lite. D. Zheleznyakov. Outline. I. Introduction II. Requirements A nd P rinciples of TBox U pdates III. Review of Model-Based Semantics IV. Review of Formula-Based Semantics V. Bold semantics VI. Conclusion. Description Logics (DLs). - PowerPoint PPT Presentation

Citation preview

Page 1: Updating TBoxes in DL-Lite

Updating TBoxes in DL-Lite

D. Zheleznyakov

Page 2: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 3: Updating TBoxes in DL-Lite

Description Logics (DLs)

Formalism to represent structered knowledge

Traditinal inference tasks for static DL KBs:– concept satisfiability– concept, role hierarchies

More recently – query answering

Web services are getting more important

Page 4: Updating TBoxes in DL-Lite

Web Services (?)

There are many things that might be called Web Services

We use the following meaning:software system designed to support interoperable machine-to-machine interaction over a network

Page 5: Updating TBoxes in DL-Lite

DLs for Web Services

Services access data through ontologies

Services can be specified using ontologies There are needs:• to enable services do data modification

ABox changes

• to modify web services

TBox changes

Page 6: Updating TBoxes in DL-Lite

Ontology Changes

There are several types of ontology changes:– Revision– Update– Smth.– and such

Page 7: Updating TBoxes in DL-Lite

Updating DL-Lite Ontologies

We study updates for DL-Lite KBs:it is the most tractable family of OWL 2

ABox updates:– Prelim./limited studied in [De Giacomo&al:2006] (?)– We revised and extended it [Calvanese&al:2010]

TBox updates:– Only TBox revision studied in [Qi,Du:2009]

– Topic of this talk is TBox updates

Page 8: Updating TBoxes in DL-Lite

Ontologies

PermStaff

Manager

AreaManager TopManager

Concepts: PermStaffManagerAreaManagerTopManager

TBox: Manager ⊑ PermStaffAreaManager ⊑ Manager

ABox: ∅

We considered TBox updates only for KBs with empty ABoxes

Page 9: Updating TBoxes in DL-Lite

Updating Ontologies

PermStaff

Manager

AreaManager TopManager

O:

U:

✓✓ ✓ ✓

Mod(O):

Page 10: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 11: Updating TBoxes in DL-Lite

Tractable Closure under Updates

We want an update operator such that: Results are expressible in DL-Lite:

we require updated KBs to be expressible in DL-Lite

Results computation is tractable:we require PTIME complexity

Page 12: Updating TBoxes in DL-Lite

Principles of TBox Updates

PermStaff

Manager

AreaManager TopManager

⊨ AreaManager PermStaff⊑TBox:

U: AreaManager ¬ PermStaff⊑

IF new TBox ⊨

AreaManager PermStaff⊑AreaManager ¬ PermStaff⊑

THEN AreaManagerM=∅ ∀M – model of the new TBox

Satisfiability Preservation:IF AM≠∅ before update,THEN AM≠∅ after update

(A is a atomic concept or role)

Page 13: Updating TBoxes in DL-Lite

Principles of TBox Updates

PermStaff

Manager

AreaManager TopManager

TBox: Manager PermSatff⊑AreaManager Manager⊑

Assume it is forbidden to changesome parts of TBox. There is a protected fragmentTpr ⊆ TBox E.g., Tpr = {Manager PermSatff⊑ }.

U: AreaManager ¬ PermStaff⊑

Protection:We accept update iffTpr and U togetherare fully satisfiable

Page 14: Updating TBoxes in DL-Lite

Principles of TBox Updates

Satisfiability Preservation:IF AM≠ ∅ before update,THEN AM≠ ∅ after update

Protection:We accept update iff protected part and U together are fully satisfiable

Moreover, we reject any update that enforces us to drop protected part

Page 15: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 16: Updating TBoxes in DL-Lite

Model-Based Semantics (MBS)PermStaff

Manager

AreaManager TopManager

O:Mod(O):

Mod(U):U:

Minimaldistance

✓ ✓ ✓ ✓

Page 17: Updating TBoxes in DL-Lite

Model-Based Semantics (MBS)PermStaff

Manager

AreaManager TopManager

Employee

Manager

AreaManager TopManager

Project?

O:

O’:

✓✓ ✓ ✓

Mod(O):

Mod(O’):

Page 18: Updating TBoxes in DL-Lite

Winslett's Semantics

What does minimal distance mean?This depends on semantics.

Winslett’s semantics:• Well known

• There are works on ABox update under Winslett’s semantics

• Representative of MBS

Page 19: Updating TBoxes in DL-Lite

Winslett's SemanticsI:

J: K:

distance(I, J) distance(I, K)

When distance(I, J) < distance(I, K) ?AI={ John, Frank }BI={ Mary }

AJ={ John }BJ={ Mary }

AK={ John }BK=∅

Page 20: Updating TBoxes in DL-Lite

Winslett's SemanticsI:

J: K:

distance(I, J) distance(I, K)

When distance(I, J) < distance(I, K) ?AI={ John, Frank }BI={ Mary }

AJ={ John }BJ={ Mary }

AK={ John }BK=∅

diff(I, J) = { {Frank}, ∅ }

Page 21: Updating TBoxes in DL-Lite

Winslett's SemanticsI:

J: K:

distance(I, J) distance(I, K)

When distance(I, J) < distance(I, K) ?AI={ John, Frank }BI={ Mary }

AJ={ John }BJ={ Mary }

AK={ John }BK=∅

diff(I, J) = { {Frank}, ∅ }diff(I, K) = { {Frank}, {Mary} }

diff(I, J) ⊂ diff(I, K)

So, distance(I, J) < distance(I, K) iff diff(I, J) ⊂ diff(I, K)

Page 22: Updating TBoxes in DL-Lite

Winslett's Semantics. Example

PermStaff

Manager

AreaManager TopManager

U: TopManager Manager⊑

What should the updated result be?

The expectation: like in the picture

Is it so under Winslett’s semantics?

Page 23: Updating TBoxes in DL-Lite

Winslett's Semantics. Example

PermStaff

Manager

AreaManager TopManager

U: TopManager Manager⊑

John

Frank

Frank

Winslett’s semantics:• new TBox ⊨ U• Mimimal change in models

John

What is a new TBox here?

new TBox: ⊨ TopManager Manager⊑ ⊨ Manager PermStuff⊑ ⊨ AreaManager Manager⊑ ⊨ AreaManager PermStaff⊑

Anything else?

✓✗??✓?✓

Page 24: Updating TBoxes in DL-Lite

Winslett's Semantics. Example

PermStaff

Manager

AreaManager TopManager

This TBox has irrelevant modelsthat cannot be obtainedfrom any model of the old TBox.

We cannot add any otherDL-Lite assertion into the new TBox,otherwise, we cut offsome relevant models

We should add something into the new TBoxto cut off them

Page 25: Updating TBoxes in DL-Lite

Winslett's Semantics

Result of update under Winslett’s semantics is inexpressible in DL-Lite.

We have to drop important assertions(Manager PermStuff)⊑

Every MBS has such a problem

Consider Formula-Based semantics

Page 26: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 27: Updating TBoxes in DL-Lite

Formula-Based Semantics (FBS)

PermStaff

Manager

AreaManager TopManager

O:

U:

PermStaff

Manager

AreaManager TopManager

Manager

AreaManager TopManager

O1:

O2:

O3:

Satisfiable

Unsatisfiable

Satisfiable

FBS: closeness is measuredbtw set of formulas

How?

We take such a subset Omax ⊆ O, which is maximal by:• cardinality, or• set inclusion, or• some preferences

Omax is not unique!There are: O1

max, O2max, …

The result is: Omax ∪ U

What to do with all of them?Depends on an approach

Page 28: Updating TBoxes in DL-Lite

WIDTIO Approach. Example

We take only those formulas that appear in every Omax:

The result is: U ∪ ∩ Ojmax

PermStaff

Manager

AreaManager TopManager

AreaManager ¬ PermStaff⊑U:TBox: AreaManager PermStaff⊑

Manager PermStaff⊑ ⊈ O1max

AreaManager Manager⊑ ⊈ O2max

j

Page 29: Updating TBoxes in DL-Lite

Cross-Product Approach. Example

OR

The output is a disjunction of KBs,one KB for each Omax:

The result is: U ∪ {∨ Ojmax}

j

AreaManager ¬ PermStaff⊑U:TBox: AreaManager PermStaff⊑

PermStaff

Manager

AreaManager TopManager

Manager PermStaff⊑ ⊈ O1max

AreaManager Manager⊑ ⊈ O2max

Page 30: Updating TBoxes in DL-Lite

Cross-Product Approach. Example

OR

The output is a disjunction of KBs,one KB for each Omax:

The result is: U ∪ {∨ Ojmax}

PermStaff

Manager

AreaManager TopManager

j

AreaManager ¬ PermStaff⊑U:TBox: AreaManager PermStaff⊑

PermStaff

Manager

AreaManager TopManager

Manager PermStaff⊑ ⊈ O1max

AreaManager Manager⊑ ⊈ O2max

Page 31: Updating TBoxes in DL-Lite

Formula-Based Semantics

WIDTIO approach:– Loses too much information

Cross-product approach:– “Keeps” too much information– Inexpressible in DL-Lite

Page 32: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 33: Updating TBoxes in DL-Lite

Bold Semantics

Which Omax to take?• A maximal one by cardinality.

NP-Hard• A maximal one by set inclusion.

Polynomial• A maximal one by some preferences

Bold approach:– Takes on board only one Omax

Page 34: Updating TBoxes in DL-Lite

Bold Semantics. Example

PermStaff

Manager

AreaManager TopManager

• Start with empty TBox

• Add assertions from U• Add assertions from TBox

one by one,if no unsatisfiability appears

AreaManager ¬ PermStaff⊑U:TBox: AreaManager Manager⊑

The result is not unique

✓✓?

Manager PermStaff⊑ ✗?AreaManager PermStaff⊑ ?✗

Page 35: Updating TBoxes in DL-Lite

Bold Semantics. Example

PermStaff

Manager

AreaManager TopManager

• Start with empty TBox

• Add assertions from U• Add assertions from TBox

one by one,if no unsatisfiability appears

AreaManager ¬ PermStaff⊑U:TBox: AreaManager Manager⊑

✓✓

Manager PermStaff⊑ ✗AreaManager PermStaff⊑ ✗

AreaManager ¬ PermStaff⊑U:TBox: AreaManager Manager⊑

✓✗

Manager PermStaff⊑ ✓AreaManager PermStaff⊑ ✗

??

?

The result is not unique

Page 36: Updating TBoxes in DL-Lite

Checking Full Satisfiability

Page 37: Updating TBoxes in DL-Lite

Outline

I. Introduction

II. Requirements And Principles ofTBox Updates

III. Review of Model-Based Semantics

IV. Review of Formula-Based Semantics

V. Bold semantics

VI. Conclusion

Page 38: Updating TBoxes in DL-Lite

Conclusion

We proposed two principles for DL KB updates

Model-based approaches:not good for TBox updates

Formula-based approaches:WIDTIO and CP are not applicableto DL-Lite KBs

Page 39: Updating TBoxes in DL-Lite

Conclusion

We proposed new semantics:Bold Semantics

We proposed polynomial time algorithmto compute update under Bold semantics

Page 40: Updating TBoxes in DL-Lite

Thank you!

Page 41: Updating TBoxes in DL-Lite

References

[De Giacomo&al:2006] [Calvanese&al:2010] [Qi,Du:2009]