Exchanging More than Complete Data
-
Upload
net2-project -
Category
Technology
-
view
180 -
download
2
description
Transcript of Exchanging More than Complete Data
![Page 1: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/1.jpg)
Exchanging More than Complete Data
Jorge PerezUniversidad de Chile
joint work with Marcelo Arenas (PUC-Chile) and Juan Reutter (Univ. Edinburgh)
![Page 2: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/2.jpg)
Father
Andy BobBob DannyDanny Eddie
Mother
Carrie Bob
Father(x , y) → Parent(x , y)Mother(x , y) → Parent(x , y)
Parent(x , y) ∧ Parent(y , z) → Gr-Parent(x , z)
![Page 3: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/3.jpg)
Father
Andy BobBob DannyDanny Eddie
Mother
Carrie Bob
Father(x , y) → Parent(x , y)Mother(x , y) → Parent(x , y)
Parent(x , y) ∧ Parent(y , z) → Gr-Parent(x , z)
SourceFather(x , y) → Pateras(x , y)
Gr-Parent(x , y) → Pappoi(x , y)Target
![Page 4: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/4.jpg)
Father
Andy BobBob DannyDanny Eddie
Mother
Carrie Bob
Father(x , y) → Parent(x , y)Mother(x , y) → Parent(x , y)
Parent(x , y) ∧ Parent(y , z) → Gr-Parent(x , z)
SourceFather(x , y) → Pateras(x , y)
Gr-Parent(x , y) → Pappoi(x , y)Target
Pateras
Andy BobBob DannyDanny Eddie
Pappoi
Andy DannyCarrie DannyBob Eddie
![Page 5: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/5.jpg)
Father
Andy BobBob DannyDanny Eddie
Mother
Carrie Bob
Father(x , y) → Parent(x , y)Mother(x , y) → Parent(x , y)
Parent(x , y) ∧ Parent(y , z) → Gr-Parent(x , z)
SourceFather(x , y) → Pateras(x , y)
Gr-Parent(x , y) → Pappoi(x , y)Target
Pateras(x , y) ∧ Pateras(y , z) → Pappoi(x , z)
![Page 6: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/6.jpg)
Father
Andy BobBob DannyDanny Eddie
Mother
Carrie Bob
Father(x , y) → Parent(x , y)Mother(x , y) → Parent(x , y)
Parent(x , y) ∧ Parent(y , z) → Gr-Parent(x , z)
SourceFather(x , y) → Pateras(x , y)
Gr-Parent(x , y) → Pappoi(x , y)Target
Pateras
Andy BobBob DannyDanny Eddie
Pappoi
Carrie Danny
Pateras(x , y) ∧ Pateras(y , z) → Pappoi(x , z)
![Page 7: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/7.jpg)
Exchanging More than Complete Data
Jorge PerezUniversidad de Chile
joint work with Marcelo Arenas (PUC-Chile) and Juan Reutter (Univ. Edinburgh)
![Page 8: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/8.jpg)
We can exchange more than complete data
◮ In data exchange we begin with a database instancewe begin with complete data
◮ What if we have a representation of a set ofpossible instances?
◮ We propose a new general formalism to exchangerepresentations of possible instancesand apply it to incomplete instances and knowledge bases
![Page 9: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/9.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 10: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/10.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 11: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/11.jpg)
Representation systems
A representation system is composed of:
◮ a set W of representatives
◮ a function rep from W to sets of instances
Vrep−→ {I1, I2, I3, . . .} = rep(V)
![Page 12: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/12.jpg)
Representation systems
A representation system is composed of:
◮ a set W of representatives
◮ a function rep from W to sets of instances
Vrep−→ {I1, I2, I3, . . .} = rep(V)
◮ Incomplete instances and knowledge bases arerepresentation systems
![Page 13: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/13.jpg)
In classical data exchange
we consider only complete data
M = (S,T,Σ) a schema mapping from S to T
I ∈ Inst(S), J ∈ Inst(T)
J is a solution for I under M iff (I , J) |= Σ
J ∈ SolM(I )
![Page 14: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/14.jpg)
In classical data exchange
we consider only complete data
M = (S,T,Σ) a schema mapping from S to T
I ∈ Inst(S), J ∈ Inst(T)
J is a solution for I under M iff (I , J) |= Σ
J ∈ SolM(I )
We can extend this to set of instances:given a set X ⊆ Inst(S)
SolM(X ) =⋃
I∈X
SolM(I )
![Page 15: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/15.jpg)
Extending the definition to representation systems
Given:
◮ a mapping M = (S,T,Σ)
◮ a representation system R = (W, rep)
◮ U ,V ∈ W
![Page 16: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/16.jpg)
Extending the definition to representation systems
Given:
◮ a mapping M = (S,T,Σ)
◮ a representation system R = (W, rep)
◮ U ,V ∈ W
Definition
U is an R-solution of V if
rep(U) ⊆ SolM(rep(V))
![Page 17: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/17.jpg)
Extending the definition to representation systems
Given:
◮ a mapping M = (S,T,Σ)
◮ a representation system R = (W, rep)
◮ U ,V ∈ W
Definition
U is an R-solution of V if
rep(U) ⊆ SolM(rep(V))
or equivalently,
∀J ∈ rep(U), ∃I ∈ rep(V) such that J ∈ SolM(I )
![Page 18: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/18.jpg)
Universal solutions and strong representation systems
Definition
U is a universal R-solution of V if
rep(U) = SolM(rep(V))
![Page 19: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/19.jpg)
Universal solutions and strong representation systems
Definition
U is a universal R-solution of V if
rep(U) = SolM(rep(V))
Let C be a class of mappings,
Definition
R = (W, rep) is a strong representation system for C iffor every M ∈ C, and V ∈ W, there exists a U ∈ W such that:
rep(U) = SolM(rep(V))
![Page 20: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/20.jpg)
Universal solutions and strong representation systems
Definition
U is a universal R-solution of V if
rep(U) = SolM(rep(V))
Let C be a class of mappings,
Definition
R = (W, rep) is a strong representation system for C iffor every M ∈ C, and V ∈ W, there exists a U ∈ W such that:
rep(U) = SolM(rep(V))
Strong representation systems ⇒ universal solutions forrepresentatives in R can always be represented in the same system.
![Page 21: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/21.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 22: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/22.jpg)
Incomplete Instances
We consider:
◮ Naive instances: instances with null values
R(1,N1)R(N1, 2)
![Page 23: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/23.jpg)
Incomplete Instances
We consider:
◮ Naive instances: instances with null values
R(1,N1)R(N1, 2)
◮ Positive conditional instances: naive instances plusconditions for tuples
◮ equalities between nulls, and between nulls and constants◮ conjunctions and disjunctions
T (1,N1) trueR(1,N1,N2) N1 = N2 ∨ (N1 = 2 ∧ N2 = 3)
![Page 24: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/24.jpg)
Incomplete Instances
We consider:
◮ Naive instances: instances with null values
R(1,N1)R(N1, 2)
◮ Positive conditional instances: naive instances plusconditions for tuples
◮ equalities between nulls, and between nulls and constants◮ conjunctions and disjunctions
T (1,N1) trueR(1,N1,N2) N1 = N2 ∨ (N1 = 2 ∧ N2 = 3)
Semantics given by valuations of nulls (+ open world):
rep(I) = {K | µ(I) ⊆ K for some valuation µ}
![Page 25: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/25.jpg)
Strong representation systems for st-tgds
Let M = (S,T,Σ) be specified by source-to-target tgds(st-tgd: φS(x) → ∃yψT(x , y), φS and ψT are CQs)
(FKMP03)
For every complete instance I there exists a naive instance J s.t.
rep(J ) = SolM(I )
![Page 26: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/26.jpg)
Strong representation systems for st-tgds
Let M = (S,T,Σ) be specified by source-to-target tgds(st-tgd: φS(x) → ∃yψT(x , y), φS and ψT are CQs)
(FKMP03)
For every complete instance I there exists a naive instance J s.t.
rep(J ) = SolM(I )
Proposition
Naive instances are not a strong representation system for st-tgds
Idea
Mng(N ,Alice) Mng(x , y) → Reports(x , y)Mng(x , x) → Self-Mng(x)
![Page 27: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/27.jpg)
Strong representation systems for st-tgds
Let M = (S,T,Σ) be specified by source-to-target tgds(st-tgd: φS(x) → ∃yψT(x , y), φS and ψT are CQs)
(FKMP03)
For every complete instance I there exists a naive instance J s.t.
rep(J ) = SolM(I )
Proposition
Naive instances are not a strong representation system for st-tgds
Idea
Mng(N ,Alice) Mng(x , y) → Reports(x , y) ?Mng(x , x) → Self-Mng(x)
![Page 28: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/28.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
![Page 29: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/29.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
Mng(N, Alice) Mng(x , y) → Reports(x , y)Mng(x , x) → Self-Mng(x)
![Page 30: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/30.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
Mng(N, Alice) Mng(x , y) → Reports(x , y) Reports(N ,Alice) true
Mng(x , x) → Self-Mng(x)
![Page 31: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/31.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
Mng(N, Alice) Mng(x , y) → Reports(x , y) Reports(N ,Alice) true
Mng(x , x) → Self-Mng(x) Self-Mng(Alice)
![Page 32: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/32.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
Mng(N, Alice) Mng(x , y) → Reports(x , y) Reports(N ,Alice) true
Mng(x , x) → Self-Mng(x) Self-Mng(Alice) N = Alice
![Page 33: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/33.jpg)
Positive conditional instances are enough
Theorem
Positive conditional instances are astrong representation system for st-tgds.
Mng(N, Alice) Mng(x , y) → Reports(x , y) Reports(N ,Alice) true
Mng(x , x) → Self-Mng(x) Self-Mng(Alice) N = Alice
Corollary
Positive conditional instances are astrong representation system for SO-tgds.
![Page 34: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/34.jpg)
Positive conditional instances are
exactly the needed representation system
Theorem
Positive conditional instances are minimal:
◮ equalities between nulls◮ equalities between constant and nulls◮ conjunctions and disjunctions
are all needed for a strong representation system of st-tgds
![Page 35: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/35.jpg)
Positive conditional instances
match the complexity bounds of classical data exchange
Theorem
For positive conditional instances and (fixed) st-tgds
◮ Universal solutions can be constructed in polynomial time.
◮ Certain answers of unions of conjunctive queriescan be computed in polynomial time.
![Page 36: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/36.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 37: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/37.jpg)
The semantics of knowledge bases
is given by sets of instances
Knowledge base over S: (I ,Γ) s.t.
I ∈ Inst(S)
Γ a set of rules over S
Semantics: finite models
Mod(I ,Γ) = {K ∈ Inst(S) | I ⊆ K and K |= Γ}
![Page 38: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/38.jpg)
We can apply our formalism
to knowledge bases
(I2,Γ2) is a kb-solution for (I1,Γ1) under M iff
Mod(I2,Γ2) ⊆ SolM(Mod(I1,Γ1))
![Page 39: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/39.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 40: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/40.jpg)
We study the following decision problem
Check-KB-Sol:
Input: M = (S,T,Σ) with Σ st-tgds(I1,Γ1) knowledge base over S with Γ1 tgds(I2,Γ2) knowledge base over T with Γ2 tgds
Output: Is (I2,Γ2) a kb-solution of (I1,Γ1) under M?
![Page 41: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/41.jpg)
We study the following decision problem
Check-KB-Sol:
Input: M = (S,T,Σ) with Σ st-tgds(I1,Γ1) knowledge base over S with Γ1 tgds(I2,Γ2) knowledge base over T with Γ2 tgds
Output: Is (I2,Γ2) a kb-solution of (I1,Γ1) under M?
Theorem
Check-KB-Sol is undecidable (even for fixed M).
![Page 42: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/42.jpg)
We study the following decision problem
Check-KB-Sol:
Input: M = (S,T,Σ) with Σ st-tgds(I1,Γ1) knowledge base over S with Γ1 tgds(I2,Γ2) knowledge base over T with Γ2 tgds
Output: Is (I2,Γ2) a kb-solution of (I1,Γ1) under M?
Theorem
Check-KB-Sol is undecidable (even for fixed M).
Undecidability is a consequence of using ∃ in knowledge bases
Check-Full-KB-Sol: Γ1, Γ2 full-tgds (no ∃)
![Page 43: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/43.jpg)
The complexity of Check-Full-KB-Sol
Theorem
Check-Full-KB-Sol is EXPTIME-complete.
![Page 44: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/44.jpg)
The complexity of Check-Full-KB-Sol
Theorem
Check-Full-KB-Sol is EXPTIME-complete.
Theorem
If M = (S,T,Σ) is fixed
Check-Full-KB-Sol is ∆P2 [O(log n)]-complete.
∆P2 [O(log n)] ≡ PNP with only a logarithmic
number of calls to the NP oracle.
![Page 45: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/45.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)
![Page 46: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/46.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
![Page 47: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/47.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
I1 and I2 are copies of G , plus a successor relation over [1, n]
![Page 48: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/48.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
I1 and I2 are copies of G , plus a successor relation over [1, n]
Γ1: “E has a clique of even size x ≤ n” → Clique(x)
![Page 49: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/49.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
I1 and I2 are copies of G , plus a successor relation over [1, n]
Γ1: “E has a clique of even size x ≤ n” → Clique(x)Γ2: “E ′ has a clique of odd size x ≤ n” → Clique′(x)
![Page 50: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/50.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
I1 and I2 are copies of G , plus a successor relation over [1, n]
Γ1: “E has a clique of even size x ≤ n” → Clique(x)Γ2: “E ′ has a clique of odd size x ≤ n” → Clique′(x)
Σ: Clique(x) ∧ Succ(x , y) → Clique′(y)
![Page 51: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/51.jpg)
Hardness is obtained by reducing
the Max-Odd-Clique problem
Given a graph G with n nodes:
S: E (·, ·), Succ(·, ·), Clique(·)T: E ′(·, ·), Succ ′(·, ·), Clique′(·)
I1 and I2 are copies of G , plus a successor relation over [1, n]
Γ1: “E has a clique of even size x ≤ n” → Clique(x)Γ2: “E ′ has a clique of odd size x ≤ n” → Clique′(x)
Σ: Clique(x) ∧ Succ(x , y) → Clique′(y)
(I2,Γ2) is a kb-solution of (I1,Σ1) iffthe maximum size of a clique in G is odd.
![Page 52: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/52.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 53: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/53.jpg)
Minimal knowledge bases as good kb-solutions
What are good knowledge-base solutions?
◮ first choice: universal kb-solutions, but
◮ there are some other kb-solutions desirable to materialize
![Page 54: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/54.jpg)
Minimal knowledge bases as good kb-solutions
What are good knowledge-base solutions?
◮ first choice: universal kb-solutions, but
◮ there are some other kb-solutions desirable to materialize
X ≡min Y: X and Y coincide in the minimal instances
![Page 55: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/55.jpg)
Minimal knowledge bases as good kb-solutions
What are good knowledge-base solutions?
◮ first choice: universal kb-solutions, but
◮ there are some other kb-solutions desirable to materialize
X ≡min Y: X and Y coincide in the minimal instances
(I2,Γ2) is a minimal kb-solution of (I1,Γ1) under M if
Mod(I2,Γ2) ≡min SolM(Mod(I1,Γ1))
We consider minimal kb-solutions as the good solutions
![Page 56: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/56.jpg)
Two requirements to construct
minimal knowledge-base solutions
Given (I1,Γ1) and Σ, when constructing aminimal-knowledge base solution (I2,Γ2) we would want:
1. Γ2 to only depend on Γ1 and Σ
Γ2 is safe for Γ1 and Σ
![Page 57: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/57.jpg)
Two requirements to construct
minimal knowledge-base solutions
Given (I1,Γ1) and Σ, when constructing aminimal-knowledge base solution (I2,Γ2) we would want:
1. Γ2 to only depend on Γ1 and Σ
Γ2 is safe for Γ1 and Σ
Definition
Γ2 is safe for Γ1 and Σ, if for every I1 there exists I2 s.t.
(I2,Γ2) is a minimal kb-solution of (I1,Γ1)
![Page 58: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/58.jpg)
Two requirements to construct
minimal knowledge-base solutions
Given (I1,Γ1) and Σ, when constructing aminimal-knowledge base solution (I2,Γ2) we would want:
2. Γ2 to be as informative as possiblethus minimizing the size of I2
![Page 59: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/59.jpg)
Two requirements to construct
minimal knowledge-base solutions
Given (I1,Γ1) and Σ, when constructing aminimal-knowledge base solution (I2,Γ2) we would want:
2. Γ2 to be as informative as possiblethus minimizing the size of I2
Γ2 is optimum-safe if for every other safe set Γ′
Γ2 |= Γ′
![Page 60: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/60.jpg)
Safe sets: data independent target knowledge bases
Unfortunately, optimum-safe sets are not (always) FO-expressible
Theorem
There exist sets Γ1 (full & acyclic) and Σ (full) s.t.
no FO sentence is optimum-safe for Γ1 and Σ
![Page 61: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/61.jpg)
Safe sets: data independent target knowledge bases
Unfortunately, optimum-safe sets are not (always) FO-expressible
Theorem
There exist sets Γ1 (full & acyclic) and Σ (full) s.t.
no FO sentence is optimum-safe for Γ1 and Σ
In our paper:
An algorithm to compute optimum-safe set in SO
![Page 62: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/62.jpg)
Safe sets: data independent target knowledge bases
Unfortunately, optimum-safe sets are not (always) FO-expressible
Theorem
There exist sets Γ1 (full & acyclic) and Σ (full) s.t.
no FO sentence is optimum-safe for Γ1 and Σ
In our paper:
An algorithm to compute optimum-safe set in SO
Can we do better? not optimum but useful set?(non-trivial safe set in a good language)
![Page 63: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/63.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
![Page 64: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/64.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
![Page 65: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/65.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
2. Construct Σ− (from T to S) as follows:
for every α(x) → R(x) in Γ+1
![Page 66: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/66.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
2. Construct Σ− (from T to S) as follows:
for every α(x) → R(x) in Γ+1
Rewrite α(x) in the target of Σ to obtain β(x)
![Page 67: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/67.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
2. Construct Σ− (from T to S) as follows:
for every α(x) → R(x) in Γ+1
Rewrite α(x) in the target of Σ to obtain β(x)
Add β(x) → R(x) to Σ−
![Page 68: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/68.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
2. Construct Σ− (from T to S) as follows:
for every α(x) → R(x) in Γ+1
Rewrite α(x) in the target of Σ to obtain β(x)
Add β(x) → R(x) to Σ−
3. Let Γ2 = Σ− ◦ Σ
![Page 69: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/69.jpg)
We can compute a good safe set
for acyclic tgds knowledge bases
Input: Γ1 acyclic full-tgds, Σ full-tgds (from S to T)
1. Unfold Γ1 to get Γ+1
2. Construct Σ− (from T to S) as follows:
for every α(x) → R(x) in Γ+1
Rewrite α(x) in the target of Σ to obtain β(x)
Add β(x) → R(x) to Σ−
3. Let Γ2 = Σ− ◦ Σ
Theorem
Γ2 is safe for Γ1 and Σ
(Γ2 is a set of full-tgds with 6=)
![Page 70: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/70.jpg)
In the paper: a complete strategy for computing
minimal knowledge-base solutions
Given Γ1 and Σ:
◮ We compute Γ2 safe for Γ1 and Σ
◮ We then compute minimal set Γ∗ (implied by Γ1) such that
for every I1
![Page 71: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/71.jpg)
In the paper: a complete strategy for computing
minimal knowledge-base solutions
Given Γ1 and Σ:
◮ We compute Γ2 safe for Γ1 and Σ
◮ We then compute minimal set Γ∗ (implied by Γ1) such that
for every I1
chaseΓ∗(I1)
![Page 72: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/72.jpg)
In the paper: a complete strategy for computing
minimal knowledge-base solutions
Given Γ1 and Σ:
◮ We compute Γ2 safe for Γ1 and Σ
◮ We then compute minimal set Γ∗ (implied by Γ1) such that
for every I1
chaseΣ( chaseΓ∗(I1))
![Page 73: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/73.jpg)
In the paper: a complete strategy for computing
minimal knowledge-base solutions
Given Γ1 and Σ:
◮ We compute Γ2 safe for Γ1 and Σ
◮ We then compute minimal set Γ∗ (implied by Γ1) such that
for every I1(
chaseΣ( chaseΓ∗(I1)),Γ2
)
![Page 74: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/74.jpg)
In the paper: a complete strategy for computing
minimal knowledge-base solutions
Given Γ1 and Σ:
◮ We compute Γ2 safe for Γ1 and Σ
◮ We then compute minimal set Γ∗ (implied by Γ1) such that
for every I1(
chaseΣ( chaseΓ∗(I1)),Γ2
)
is a minimal kb-solution for (I1,Γ1).
![Page 75: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/75.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks
![Page 76: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/76.jpg)
We can exchange more than complete data
We propose a general formalismto exchange representation systems
◮ Applications to incomplete instances
◮ Applications to knowledge bases
![Page 77: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/77.jpg)
We can exchange more than complete data
We propose a general formalismto exchange representation systems
◮ Applications to incomplete instances
◮ Applications to knowledge bases
Next step: apply our general setting to the Semantic Web
◮ Semantic Web data have nulls (blank nodes)
◮ Semantic Web specifications have rules (RDFS, OWL)
![Page 78: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/78.jpg)
Exchanging More than Complete Data
Jorge PerezUniversidad de Chile
joint work with Marcelo Arenas (PUC-Chile) and Juan Reutter (Univ. Edinburgh)
![Page 79: Exchanging More than Complete Data](https://reader035.fdocuments.net/reader035/viewer/2022062616/5490b2c8b4795976328b46ce/html5/thumbnails/79.jpg)
Outline
Formalism for exchanging representations systems
Applications to incomplete instances
Applications to knowledge basesWhat is the complexity of testing kb-solutions?How can we compute a good kb-solution?
Concluding remarks