Federation and Navigation in SPARQL 1.1
Jorge P´erez
Assistant Professor
Department of Computer Science
Universidad de Chi...
Outline
Basics of SPARQL
Syntax and Semantics of SPARQL 1.0
What is new in SPARQL 1.1
Federation: SERVICE operator
Syntax ...
SPARQL query language for RDF
SPARQL query language for RDF
RDF Graph:
fmeza@utalca.cl
:name
:email
:phone
:name
:friendOf rgarrido@utalca.cl:email
Fede...
SPARQL query language for RDF
RDF Graph:
fmeza@utalca.cl
:name
:email
:phone
:name
:friendOf rgarrido@utalca.cl:email
Fede...
SPARQL query language for RDF
RDF Graph:
fmeza@utalca.cl
:name
:email
:phone
:name
:friendOf rgarrido@utalca.cl:email
Fede...
SPARQL query language for RDF
RDF Graph:
fmeza@utalca.cl
:name
:email
:phone
:name
:friendOf rgarrido@utalca.cl:email
Fede...
An example of an RDF graph to query: DBLP
inAMW:SarmaUW09 :Jeffrey D. Ullman
:Anish Das Sarma
:Jennifer Widom
inAMW:2009
"...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
}
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC
SELECT ?Author
WHERE
{
?Paper dc:creator ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC, and their Web
pages if this information ...
SPARQL: A Simple RDF Query Language
Example: Authors that have published in ISWC, and their Web
pages if this information ...
But things can become more complex...
Interesting features of pattern
matching on graphs SELECT ?X1 ?X2 ...
{ P1 .
P2 }
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
SELECT ?X1 ?X2 ...
{{ ...
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
◮ Optional parts
SELEC...
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
◮ Optional parts
◮ Nes...
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
◮ Optional parts
◮ Nes...
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
◮ Optional parts
◮ Nes...
But things can become more complex...
Interesting features of pattern
matching on graphs
◮ Grouping
◮ Optional parts
◮ Nes...
Outline
Basics of SPARQL
Syntax and Semantics of SPARQL 1.0
What is new in SPARQL 1.1
Federation: SERVICE operator
Syntax ...
RDF triples and graphs
Subject Object
Predicate
LB
U
U UB
U : set of URIs
B : set of blank nodes
L : set of literals
RDF triples and graphs
Subject Object
Predicate
LB
U
U UB
U : set of URIs
B : set of blank nodes
L : set of literals
(s, p...
RDF triples and graphs
Subject Object
Predicate
LB
U
U UB
U : set of URIs
B : set of blank nodes
L : set of literals
(s, p...
RDF triples and graphs
Subject Object
Predicate
LB
U
U UB
U : set of URIs
B : set of blank nodes
L : set of literals
(s, p...
A standard algebraic syntax
◮ Triple patterns: just RDF triples + variables (from a set V )
?X :name "john" (?X, name, joh...
A standard algebraic syntax
◮ Triple patterns: just RDF triples + variables (from a set V )
?X :name "john" (?X, name, joh...
A standard algebraic syntax
◮ Triple patterns: just RDF triples + variables (from a set V )
?X :name "john" (?X, name, joh...
A standard algebraic syntax
◮ Triple patterns: just RDF triples + variables (from a set V )
?X :name "john" (?X, name, joh...
A standard algebraic syntax
◮ Triple patterns: just RDF triples + variables (from a set V )
?X :name "john" (?X, name, joh...
A standard algebraic syntax (cont.)
◮ Explicit precedence/association
Example
{ t1
t2
OPTIONAL { t3 }
OPTIONAL { t4 }
t5
}...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
Mappings: building block for the semantics
Definition
A mapping is a partial function from variables to RDF terms
µ : V → U...
The semantics of triple patterns
Definition
The evaluation of triple patter t over a graph G, denoted by t G ,
is the set o...
The semantics of triple patterns
Definition
The evaluation of triple patter t over a graph G, denoted by t G ,
is the set o...
The semantics of triple patterns
Definition
The evaluation of triple patter t over a graph G, denoted by t G ,
is the set o...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(?X, name, ?N) G
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(?X, name, ?N) G
µ1 = {?X → R1, ?N → john}
µ2 = {?X → R2,...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(?X, name, ?N) G
µ1 = {?X → R1, ?N → john}
µ2 = {?X → R2,...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(?X, name, ?N) G
µ1 = {?X → R1, ?N → john}
µ2 = {?X → R2,...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(?X, name, ?N) G
?X ?N
µ1 R1 john
µ2 R2 paul
(?X, email, ...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(R1, webPage, ?W ) G
(R2, name, paul) G
(R3, name, ringo)...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(R1, webPage, ?W ) G
(R2, name, paul) G
(R3, name, ringo)...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(R1, webPage, ?W ) G
(R2, name, paul) G
(R3, name, ringo)...
Example
G
(R1, name, john)
(R1, email, J@ed.ex)
(R2, name, paul)
(R1, webPage, ?W ) G
(R2, name, paul) G
µ∅ = { }
(R3, nam...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Compatible mappings: mappings that can be merged.
Definition
The mappings µ1, µ2 are compatibles iff they agree
in their sha...
Sets of mappings and operations
Let M1 and M2 be sets of mappings:
Definition
Join: M1 ⋊⋉ M2
◮ {µ1 ∪ µ2 | µ1 ∈ M1, µ2 ∈ M2,...
Sets of mappings and operations
Let M1 and M2 be sets of mappings:
Definition
Join: M1 ⋊⋉ M2
◮ {µ1 ∪ µ2 | µ1 ∈ M1, µ2 ∈ M2,...
Sets of mappings and operations
Definition
Difference: M1 M2
◮ {µ ∈ M1 | for all µ′ ∈ M2, µ and µ′ are not compatibles}
◮ ma...
Sets of mappings and operations
Definition
Difference: M1 M2
◮ {µ ∈ M1 | for all µ′ ∈ M2, µ and µ′ are not compatibles}
◮ ma...
Semantics of general graph patterns
Definition
Given a graph G the evaluation of a pattern is recursively defined
the base c...
Semantics of general graph patterns
Definition
Given a graph G the evaluation of a pattern is recursively defined
◮ (P1 AND ...
Semantics of general graph patterns
Definition
Given a graph G the evaluation of a pattern is recursively defined
◮ (P1 AND ...
Semantics of general graph patterns
Definition
Given a graph G the evaluation of a pattern is recursively defined
◮ (P1 AND ...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (AND)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (OPT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, webPa...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Example (UNION)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, web...
Boolean filter expressions (value constraints)
In filter expressions we consider
◮ the equality = among variables and RDF te...
Satisfaction of value constraints
◮ If P is a graph pattern and R is a value constraint then
(P FILTER R) is also a graph ...
Satisfaction of value constraints
◮ If P is a graph pattern and R is a value constraint then
(P FILTER R) is also a graph ...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (FILTER)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Why do we need/want to formalize SPARQL
A formalization is beneficial
◮ clarifying corner cases
◮ helping in the implementa...
The evaluation decision problem
Evaluation problem for SPARQL patterns
Input: A mapping µ, an RDF graph G
a graph pattern ...
Brief complexity-theory reminder
P
Brief complexity-theory reminder
NP
P
Brief complexity-theory reminder
PSPACE
NP
P
Complexity of the evaluation problem
Theorem (PAG09)
For patterns using only the AND operator,
the evaluation problem is i...
Complexity of the evaluation problem
Theorem (PAG09)
For patterns using only the AND operator,
the evaluation problem is i...
Complexity of the evaluation problem
Theorem (PAG09)
For patterns using only the AND operator,
the evaluation problem is i...
Complexity of the evaluation problem
Theorem (PAG09)
For patterns using only the AND operator,
the evaluation problem is i...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
Well–designed patterns
Can we find a natural fragment with better complexity?
Definition
An AND-OPT pattern is well–designed...
A bit more on complexity...
PSPACE
NP
P
A bit more on complexity...
PSPACE
coNPNP
P
Evaluation of well-designed patterns is in coNP-complete
Theorem (PAG09)
For AND-OPT well–designed graph patterns
the eval...
Evaluation of well-designed patterns is in coNP-complete
Theorem (PAG09)
For AND-OPT well–designed graph patterns
the eval...
SELECT (a.k.a. projection)
Besides graph patterns, SPARQL 1.0 allow result forms
the most simple is SELECT
Definition
A SEL...
SELECT (a.k.a. projection)
Besides graph patterns, SPARQL 1.0 allow result forms
the most simple is SELECT
Definition
A SEL...
Example (SELECT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (SELECT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
Example (SELECT)
G :
(R1, name, john) (R2, name, paul) (R3, name, ringo)
(R1, email, J@ed.ex) (R3, email, R@ed.ex)
(R3, we...
SPARQL 1.1 introduces several new features
In SPARQL 1.1:
◮ (SELECT W P) can be used as any other graph pattern
(ASK P) ca...
Outline
Basics of SPARQL
Syntax and Semantics of SPARQL 1.0
What is new in SPARQL 1.1
Federation: SERVICE operator
Syntax ...
SPARQL endpoints and the SERVICE operator
◮ SPARQL endpoints are services that accepts HTTP requests
asking for SPARQL que...
SPARQL endpoints and the SERVICE operator
◮ SPARQL endpoints are services that accepts HTTP requests
asking for SPARQL que...
SPARQL endpoints and the SERVICE operator
◮ SPARQL endpoints are services that accepts HTTP requests
asking for SPARQL que...
Syntax of SERVICE
Syntax
If c ∈ U, ?X is a variable, and P is a graph pattern then
Syntax of SERVICE
Syntax
If c ∈ U, ?X is a variable, and P is a graph pattern then
◮ (SERVICE c P)
◮ (SERVICE ?X P)
are SE...
Syntax of SERVICE
Syntax
If c ∈ U, ?X is a variable, and P is a graph pattern then
◮ (SERVICE c P)
◮ (SERVICE ?X P)
are SE...
Semantics of SERVICE
We assume the existence of a partial function
ep : U → RDF graphs
intuitively, ep(c) is the (default)...
Semantics of SERVICE
We assume the existence of a partial function
ep : U → RDF graphs
intuitively, ep(c) is the (default)...
Semantics of SERVICE
We assume the existence of a partial function
ep : U → RDF graphs
intuitively, ep(c) is the (default)...
Semantics of SERVICE
We assume the existence of a partial function
ep : U → RDF graphs
intuitively, ep(c) is the (default)...
Semantics of SERVICE
We assume the existence of a partial function
ep : U → RDF graphs
intuitively, ep(c) is the (default)...
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G =
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G = P ep(c)
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G = P ep(c) ⋊⋉ {{...
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G =
c∈dom(ep)
P e...
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G =
c∈dom(ep)
P e...
Semantics of SERVICE
Definition
Given an RDF graph G, a variable ?X, and a graph pattern P
(SERVICE ?X P) G =
c∈dom(ep)
P e...
SERVICE example
Some queries/patterns can be safely evaluated:
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
SERVICE example
Some queries/patterns can be safely evaluated:
◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?...
Unbounded SERVICE queries
What about this pattern?
(?X, service description, ?Z) UNION (?X, service address, ?Y )
AND (SER...
Unbounded SERVICE queries
What about this pattern?
(?X, service description, ?Z) UNION (?X, service address, ?Y )
AND (SER...
Unbounded SERVICE queries
What about this pattern?
(?X, service description, ?Z) UNION (?X, service address, ?Y )
AND (SER...
Boundedness
Definition
A variable ?X is bound in graph pattern P if
for every graph G and every µ ∈ P G it holds that:
◮ ?X...
Boundedness
Definition
A variable ?X is bound in graph pattern P if
for every graph G and every µ ∈ P G it holds that:
◮ ?X...
Boundedness
Definition
A variable ?X is bound in graph pattern P if
for every graph G and every µ ∈ P G it holds that:
◮ ?X...
Boundedness
Definition
A variable ?X is bound in graph pattern P if
for every graph G and every µ ∈ P G it holds that:
◮ ?X...
Boundedness
Definition
A variable ?X is bound in graph pattern P if
for every graph G and every µ ∈ P G it holds that:
◮ ?X...
Undecidability of boundedness
Proof idea
◮ From [AG08]: it is undecidable to check if a SPARQL pattern
P is satisfiable (if...
Undecidability of boundedness
Proof idea
◮ From [AG08]: it is undecidable to check if a SPARQL pattern
P is satisfiable (if...
Undecidability of boundedness
Proof idea
◮ From [AG08]: it is undecidable to check if a SPARQL pattern
P is satisfiable (if...
Undecidability of boundedness
Proof idea
◮ From [AG08]: it is undecidable to check if a SPARQL pattern
P is satisfiable (if...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
A sufficient condition for boundedness
Definition [BAC11]
The set of strongly bounded variables in a pattern P, denoted by
SB...
(Strongly) boundedness is not enough
Are we happy now?
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y )...
(Strongly) boundedness is not enough
Are we happy now?
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y )...
(Strongly) boundedness is not enough
Are we happy now?
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y )...
(Strongly) boundedness is not enough
Are we happy now?
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y )...
(Strongly) boundedness is not enough
Are we happy now?
(Strongly) boundedness is not enough
Are we happy now?
P2 = (?U1, related with, ?U2) AND
SERVICE ?U1 (?N, email, ?E) OPT
(...
(Strongly) boundedness is not enough
Are we happy now?
P2 = (?U1, related with, ?U2) AND
SERVICE ?U1 (?N, email, ?E) OPT
(...
(Strongly) boundedness is not enough
Are we happy now?
P2 = (?U1, related with, ?U2) AND
SERVICE ?U1 (?N, email, ?E) OPT
(...
(Strongly) boundedness is not enough
Are we happy now?
P2 = (?U1, related with, ?U2) AND
SERVICE ?U1 (?N, email, ?E) OPT
(...
Parse tree of a pattern
We need first to formalize the tree of subexpressions of a pattern
(?Y , a, ?Z) UNION (?X, b, c) AN...
Parse tree of a pattern
We need first to formalize the tree of subexpressions of a pattern
(?Y , a, ?Z) UNION (?X, b, c) AN...
Parse tree of a pattern
We need first to formalize the tree of subexpressions of a pattern
(?Y , a, ?Z) UNION (?X, b, c) AN...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-boundedness
Notion of boundedness considering the important part
Definition
A pattern P is service-bound if for eve...
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Definition
A ...
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Definition
A ...
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Definition
A ...
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Definition
A ...
Service-safeness
We need a decidable sufficient condition.
Idea: replace bound by SB(·) in the previous notion.
Definition
A ...
Applying service-safeness
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y ) AND
(SERVICE ?Y (?N, email, ...
Applying service-safeness
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y ) AND
(SERVICE ?Y (?N, email, ...
Applying service-safeness
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y ) AND
(SERVICE ?Y (?N, email, ...
Applying service-safeness
P1 = (?X, service description, ?Z) UNION
(?X, service address, ?Y ) AND
(SERVICE ?Y (?N, email, ...
Closing words on SERVICE
SERVICE: several interesting research question
◮ Optimization (rewriting, reordering, containment...
Outline
Basics of SPARQL
Syntax and Semantics of SPARQL 1.0
What is new in SPARQL 1.1
Federation: SERVICE operator
Syntax ...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.0 has very limited navigational capabilities
Assume a graph with cities and connections with RDF triples like:
(C...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
SPARQL 1.1 provides an alternative way for navigating
URI2
:email
seba@puc.cl
:name
:email
:phone
:name
:friendOf juan@ute...
General navigation using regular expressions
Regular expressions define sets of strings using
◮ concatenation: /
◮ disjunct...
General navigation using regular expressions
Regular expressions define sets of strings using
◮ concatenation: /
◮ disjunct...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
Interesting navigational queries
◮ RDF graph with :father, :mother edges:
ancestors of John
{ John (:father|:mother)* ?X }...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
As always, we need a (formal) semantics
◮ Regular expressions in SPARQL queries seem reasonable
◮ We need to agree in the ...
SPARQL 1.1 implementations
had a poor performance
Data:
◮ cliques (complete graphs) of different size
◮ from 2 nodes (87 by...
SPARQL 1.1 implementations
had a poor performance
1
10
100
1000
2 4 6 8 10 12 14 16
ARQ
+ + + + + + +
+
+
+
+
+
RDFQ
× × ×...
Poor performance with real Web data of small size
Data:
◮ Social Network data given by foaf:knows links
◮ Crawled from Axe...
Poor performance with real Web data of small size
SELECT * WHERE { axel:me (foaf:knows)* ?x }
Poor performance with real Web data of small size
SELECT * WHERE { axel:me (foaf:knows)* ?x }
Input ARQ RDFQ Kgram Sesame
...
Poor performance with real Web data of small size
SELECT * WHERE { axel:me (foaf:knows)* ?x }
Input ARQ RDFQ Kgram Sesame
...
This is a problem of the specification
[ACP12]
Any implementation that follows SPARQL 1.1 standard
(as of January 2012) is ...
This is a problem of the specification
[ACP12]
Any implementation that follows SPARQL 1.1 standard
(as of January 2012) is ...
This is a problem of the specification
[ACP12]
Any implementation that follows SPARQL 1.1 standard
(as of January 2012) is ...
SPARQL 1.1 property paths match regular expressions
but also count
:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X }
SPARQL 1.1 property paths match regular expressions
but also count
:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X }...
SPARQL 1.1 property paths match regular expressions
but also count
:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X }...
SPARQL 1.1 property paths match regular expressions
but also count
:p:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X...
SPARQL 1.1 property paths match regular expressions
but also count
:p:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X...
SPARQL 1.1 property paths match regular expressions
but also count
:p:a
:b
:c
:p
:p
:p
:p
:d
SELECT ?X
WHERE { :a (:p)* ?X...
SPARQL 1.1 document provides a special procedure
to handle cycles (and make the count)
Evaluation of path*
“the algorithm ...
SPARQL 1.1 document provides a special procedure
to handle cycles (and make the count)
Evaluation of path*
“the algorithm ...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 }
every solution is a copy of the empty mapping...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 }
n # Sol.
9 13,700
10 109,601
11 986,410
12 9,...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 }
n # Sol.
9 13,700
10 109...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 }
n # Sol.
9 13,700
10 109...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } { :a0 (((:p)*)*)* :a1 }
...
Counting the number of solutions...
Data: Clique of size n
{ :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } { :a0 (((:p)*)*)* :a1 }
...
More on counting the number of solutions...
Data: foaf links crawled from the Web
{ axel:me (foaf:knows)* ?x }
More on counting the number of solutions...
Data: foaf links crawled from the Web
{ axel:me (foaf:knows)* ?x }
File # URIs...
More on counting the number of solutions...
Data: foaf links crawled from the Web
{ axel:me (foaf:knows)* ?x }
File # URIs...
More on counting the number of solutions...
Data: foaf links crawled from the Web
{ axel:me (foaf:knows)* ?x }
File # URIs...
A bit more on complexity classes...
Complexity can be measured by using counting-complexity classes
NP #P
Sat: is a propos...
A bit more on complexity classes...
Complexity can be measured by using counting-complexity classes
NP #P
Sat: is a propos...
A bit more on complexity classes...
Complexity can be measured by using counting-complexity classes
NP #P
Sat: is a propos...
Counting problem for property paths
CountW3C
Input: RDF graph G
Property path triple { :a path :b }
Output: Count the numb...
The complexity of property paths is intractable
Theorem (ACP12)
CountW3C is outside #P
The complexity of property paths is intractable
Theorem (ACP12)
CountW3C is outside #P
CountW3C is hard to solve even if P...
A doubly exponential lower bound for counting
◮ Let paths be a property path of the form
(· · · ((:p)*)*)· · ·)*
with s ne...
A doubly exponential lower bound for counting
◮ Let paths be a property path of the form
(· · · ((:p)*)*)· · ·)*
with s ne...
A doubly exponential lower bound for counting
◮ Let paths be a property path of the form
(· · · ((:p)*)*)· · ·)*
with s ne...
A doubly exponential lower bound for counting
◮ Let paths be a property path of the form
(· · · ((:p)*)*)· · ·)*
with s ne...
A doubly exponential lower bound for counting
◮ Let paths be a property path of the form
(· · · ((:p)*)*)· · ·)*
with s ne...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
We can explain the experimental results
CountClique(s, n) allows to fill in the blanks
{ :a0 ((:p)*)* :a1 }
n # Sol.
2 1
3 ...
Data complexity of property path is still intractable
Common assumption in Databases:
◮ queries are much smaller than data...
Data complexity of property path is still intractable
Common assumption in Databases:
◮ queries are much smaller than data...
Data complexity of property path is still intractable
Common assumption in Databases:
◮ queries are much smaller than data...
Data complexity of property path is still intractable
Common assumption in Databases:
◮ queries are much smaller than data...
Data complexity of property path is still intractable
Common assumption in Databases:
◮ queries are much smaller than data...
Existential semantics: a possible alternative
Possible solution
Do not count
Just check whether there exists a path
satisf...
Existential semantics: a possible alternative
Possible solution
Do not count
Just check whether there exists a path
satisf...
Existential semantics: decision problems
Input: RDF graph G
Property path triple { :a path :b }
ExistsPath
Question: Is th...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Evaluating existential paths is tractable
Theorem (well-known result)
ExistsPath can be solved in O(|G| × |path|)
Can be p...
Relationship between ExistsPath and ExistsW3C
Theorem (ACP12)
ExistsPath and ExistsW3C are equivalent decision problems
Relationship between ExistsPath and ExistsW3C
Theorem (ACP12)
ExistsPath and ExistsW3C are equivalent decision problems
Co...
So there are possibilities for optimization
SPARQL includes an operator to eliminate duplicates (DISTINCT)
So there are possibilities for optimization
SPARQL includes an operator to eliminate duplicates (DISTINCT)
Corollary
Prope...
So there are possibilities for optimization
SPARQL includes an operator to eliminate duplicates (DISTINCT)
Corollary
Prope...
SPARQL 1.1 implementations
do not take advantage of SELECT DISTINCT
SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x }
I...
SPARQL 1.1 implementations
do not take advantage of SELECT DISTINCT
SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x }
I...
SPARQL 1.1 implementations
do not take advantage of SELECT DISTINCT
SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x }
I...
SPARQL 1.1 implementations
do not take advantage of SELECT DISTINCT
SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x }
I...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
New semantics for property paths (July 2012)
◮ Paths constructed only from / and | should be counted
◮ As soon as * is use...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Is the new semantics the right one?
W3C definitely wants to count paths.
Are there more reasonable alternatives?
◮ to have ...
Several work to do on navigational queries for graphs!
Other lines of research with open questions:
◮ Nested regular expre...
Outline
Basics of SPARQL
Syntax and Semantics of SPARQL 1.0
What is new in SPARQL 1.1
Federation: SERVICE operator
Syntax ...
Concluding remarks
◮ Federation and Navigation are fundamental features in
SPARQL 1.1
◮ They need formalization and (serio...
Concluding remarks
◮ Federation and Navigation are fundamental features in
SPARQL 1.1
◮ They need formalization and (serio...
Federation and Navigation in SPARQL 1.1
Federation and Navigation in SPARQL 1.1
Federation and Navigation in SPARQL 1.1
Federation and Navigation in SPARQL 1.1
Upcoming SlideShare
Loading in...5
×

Federation and Navigation in SPARQL 1.1

356
-1

Published on

Published in: Education
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
356
On Slideshare
0
From Embeds
0
Number of Embeds
1
Actions
Shares
0
Downloads
9
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

Federation and Navigation in SPARQL 1.1

  1. 1. Federation and Navigation in SPARQL 1.1 Jorge P´erez Assistant Professor Department of Computer Science Universidad de Chile
  2. 2. Outline Basics of SPARQL Syntax and Semantics of SPARQL 1.0 What is new in SPARQL 1.1 Federation: SERVICE operator Syntax and Semantics Evaluation of SERVICE queries Navigation: Property Paths Navigating graphs with regular expressions The history of paths (in SPARQL 1.1 specification) Evaluation procedures and complexity
  3. 3. SPARQL query language for RDF
  4. 4. SPARQL query language for RDF RDF Graph: fmeza@utalca.cl :name :email :phone :name :friendOf rgarrido@utalca.cl:email Federico Meza 35-446928 Ruth Garrido URI2 URI1 RDF-triples: (URI2, :email, rgarrido@utalca.cl)
  5. 5. SPARQL query language for RDF RDF Graph: fmeza@utalca.cl :name :email :phone :name :friendOf rgarrido@utalca.cl:email Federico Meza 35-446928 Ruth Garrido URI2 URI1 RDF-triples: (URI2, :email, rgarrido@utalca.cl) SPARQL Query: SELECT ?N WHERE { ?X :name ?N . }
  6. 6. SPARQL query language for RDF RDF Graph: fmeza@utalca.cl :name :email :phone :name :friendOf rgarrido@utalca.cl:email Federico Meza 35-446928 Ruth Garrido URI2 URI1 RDF-triples: (URI2, :email, rgarrido@utalca.cl) SPARQL Query: SELECT ?N ?E WHERE { ?X :name ?N . ?X :email ?E . }
  7. 7. SPARQL query language for RDF RDF Graph: fmeza@utalca.cl :name :email :phone :name :friendOf rgarrido@utalca.cl:email Federico Meza 35-446928 Ruth Garrido URI2 URI1 RDF-triples: (URI2, :email, rgarrido@utalca.cl) SPARQL Query: SELECT ?N ?E WHERE { ?X :name ?N . ?X :email ?E . ?X :friendOf ?Y . ?Y :name "Ruth Garrido" }
  8. 8. An example of an RDF graph to query: DBLP inAMW:SarmaUW09 :Jeffrey D. Ullman :Anish Das Sarma :Jennifer Widom inAMW:2009 "Schema Design for ..." dc:creator dc:creator dc:creator dct:partOf dc:title swrc:series conf:amw <http://purl.org/dc/terms/> : <http://dblp.l3s.de/d2r/resource/authors/> conf: <http://dblp.l3s.de/d2r/resource/conferences/> inAMW: <http://dblp.l3s.de/d2r/resource/publications/conf/amw/> swrc: <http://swrc.ontoware.org/ontology#> dc: dct: <http://purl.org/dc/elements/1.1/>
  9. 9. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC
  10. 10. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author
  11. 11. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { }
  12. 12. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . }
  13. 13. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . }
  14. 14. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . }
  15. 15. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . } A SPARQL query consists of a:
  16. 16. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . } A SPARQL query consists of a: Head: Processing of the variables
  17. 17. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . } A SPARQL query consists of a: Head: Processing of the variables Body: Pattern matching expression
  18. 18. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC, and their Web pages if this information is available: SELECT ?Author ?WebPage WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . OPTIONAL { ?Author foaf:homePage ?WebPage . } }
  19. 19. SPARQL: A Simple RDF Query Language Example: Authors that have published in ISWC, and their Web pages if this information is available: SELECT ?Author ?WebPage WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . OPTIONAL { ?Author foaf:homePage ?WebPage . } }
  20. 20. But things can become more complex... Interesting features of pattern matching on graphs SELECT ?X1 ?X2 ... { P1 . P2 }
  21. 21. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping SELECT ?X1 ?X2 ... {{ P1 . P2 } { P3 . P4 } }
  22. 22. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping ◮ Optional parts SELECT ?X1 ?X2 ... {{ P1 . P2 OPTIONAL { P5 } } { P3 . P4 OPTIONAL { P7 } } }
  23. 23. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping ◮ Optional parts ◮ Nesting SELECT ?X1 ?X2 ... {{ P1 . P2 OPTIONAL { P5 } } { P3 . P4 OPTIONAL { P7 OPTIONAL { P8 } } } }
  24. 24. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping ◮ Optional parts ◮ Nesting ◮ Union of patterns SELECT ?X1 ?X2 ... {{{ P1 . P2 OPTIONAL { P5 } } { P3 . P4 OPTIONAL { P7 OPTIONAL { P8 } } } } UNION { P9 }}
  25. 25. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping ◮ Optional parts ◮ Nesting ◮ Union of patterns ◮ Filtering ◮ ... ◮ + several new features in the upcoming version: federation, navigation SELECT ?X1 ?X2 ... {{{ P1 . P2 OPTIONAL { P5 } } { P3 . P4 OPTIONAL { P7 OPTIONAL { P8 } } } } UNION { P9 FILTER ( R ) }}
  26. 26. But things can become more complex... Interesting features of pattern matching on graphs ◮ Grouping ◮ Optional parts ◮ Nesting ◮ Union of patterns ◮ Filtering ◮ ... ◮ + several new features in the upcoming version: federation, navigation SELECT ?X1 ?X2 ... {{{ P1 . P2 OPTIONAL { P5 } } { P3 . P4 OPTIONAL { P7 OPTIONAL { P8 } } } } UNION { P9 FILTER ( R ) }} What is the (formal) meaning of a general SPARQL query?
  27. 27. Outline Basics of SPARQL Syntax and Semantics of SPARQL 1.0 What is new in SPARQL 1.1 Federation: SERVICE operator Syntax and Semantics Evaluation of SERVICE queries Navigation: Property Paths Navigating graphs with regular expressions The history of paths (in SPARQL 1.1 specification) Evaluation procedures and complexity
  28. 28. RDF triples and graphs Subject Object Predicate LB U U UB U : set of URIs B : set of blank nodes L : set of literals
  29. 29. RDF triples and graphs Subject Object Predicate LB U U UB U : set of URIs B : set of blank nodes L : set of literals (s, p, o) ∈ (U ∪ B) × U × (U ∪ B ∪ L) is called an RDF triple
  30. 30. RDF triples and graphs Subject Object Predicate LB U U UB U : set of URIs B : set of blank nodes L : set of literals (s, p, o) ∈ (U ∪ B) × U × (U ∪ B ∪ L) is called an RDF triple A set of RDF triples is called an RDF graph
  31. 31. RDF triples and graphs Subject Object Predicate LB U U UB U : set of URIs B : set of blank nodes L : set of literals (s, p, o) ∈ (U ∪ B) × U × (U ∪ B ∪ L) is called an RDF triple A set of RDF triples is called an RDF graph In this talk, we do not consider blank nodes ◮ (s, p, o) ∈ U × U × (U ∪ L) is called an RDF triple
  32. 32. A standard algebraic syntax ◮ Triple patterns: just RDF triples + variables (from a set V ) ?X :name "john" (?X, name, john)
  33. 33. A standard algebraic syntax ◮ Triple patterns: just RDF triples + variables (from a set V ) ?X :name "john" (?X, name, john) ◮ Graph patterns: full parenthesized algebra original SPARQL syntax algebraic syntax { P1 . P2 } ( P1 AND P2 )
  34. 34. A standard algebraic syntax ◮ Triple patterns: just RDF triples + variables (from a set V ) ?X :name "john" (?X, name, john) ◮ Graph patterns: full parenthesized algebra original SPARQL syntax algebraic syntax { P1 . P2 } ( P1 AND P2 ) { P1 OPTIONAL { P2 }} ( P1 OPT P2 )
  35. 35. A standard algebraic syntax ◮ Triple patterns: just RDF triples + variables (from a set V ) ?X :name "john" (?X, name, john) ◮ Graph patterns: full parenthesized algebra original SPARQL syntax algebraic syntax { P1 . P2 } ( P1 AND P2 ) { P1 OPTIONAL { P2 }} ( P1 OPT P2 ) { P1 } UNION { P2 } ( P1 UNION P2 )
  36. 36. A standard algebraic syntax ◮ Triple patterns: just RDF triples + variables (from a set V ) ?X :name "john" (?X, name, john) ◮ Graph patterns: full parenthesized algebra original SPARQL syntax algebraic syntax { P1 . P2 } ( P1 AND P2 ) { P1 OPTIONAL { P2 }} ( P1 OPT P2 ) { P1 } UNION { P2 } ( P1 UNION P2 ) { P1 FILTER ( R ) } ( P1 FILTER R )
  37. 37. A standard algebraic syntax (cont.) ◮ Explicit precedence/association Example { t1 t2 OPTIONAL { t3 } OPTIONAL { t4 } t5 } ( ( ( ( t1 AND t2 ) OPT t3 ) OPT t4 ) AND t5 )
  38. 38. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L
  39. 39. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t:
  40. 40. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t: ◮ µ(t): triple obtained from t replacing variables according to µ
  41. 41. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t: ◮ µ(t): triple obtained from t replacing variables according to µ Example
  42. 42. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t: ◮ µ(t): triple obtained from t replacing variables according to µ Example µ = {?X → R1, ?Y → R2, ?Name → john}
  43. 43. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t: ◮ µ(t): triple obtained from t replacing variables according to µ Example µ = {?X → R1, ?Y → R2, ?Name → john} t = (?X, name, ?Name)
  44. 44. Mappings: building block for the semantics Definition A mapping is a partial function from variables to RDF terms µ : V → U ∪ L Given a mapping µ and a triple pattern t: ◮ µ(t): triple obtained from t replacing variables according to µ Example µ = {?X → R1, ?Y → R2, ?Name → john} t = (?X, name, ?Name) µ(t) = (R1, name, john)
  45. 45. The semantics of triple patterns Definition The evaluation of triple patter t over a graph G, denoted by t G , is the set of all mappings µ such that:
  46. 46. The semantics of triple patterns Definition The evaluation of triple patter t over a graph G, denoted by t G , is the set of all mappings µ such that: ◮ dom(µ) is exactly the set of variables occurring in t
  47. 47. The semantics of triple patterns Definition The evaluation of triple patter t over a graph G, denoted by t G , is the set of all mappings µ such that: ◮ dom(µ) is exactly the set of variables occurring in t ◮ µ(t) ∈ G
  48. 48. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (?X, name, ?N) G
  49. 49. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (?X, name, ?N) G µ1 = {?X → R1, ?N → john} µ2 = {?X → R2, ?N → paul}
  50. 50. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (?X, name, ?N) G µ1 = {?X → R1, ?N → john} µ2 = {?X → R2, ?N → paul} (?X, email, ?E) G
  51. 51. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (?X, name, ?N) G µ1 = {?X → R1, ?N → john} µ2 = {?X → R2, ?N → paul} (?X, email, ?E) G µ = {?X → R1, ?E → J@ed.ex}
  52. 52. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (?X, name, ?N) G ?X ?N µ1 R1 john µ2 R2 paul (?X, email, ?E) G ?X ?E µ R1 J@ed.ex
  53. 53. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (R1, webPage, ?W ) G (R2, name, paul) G (R3, name, ringo) G
  54. 54. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (R1, webPage, ?W ) G (R2, name, paul) G (R3, name, ringo) G
  55. 55. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (R1, webPage, ?W ) G (R2, name, paul) G (R3, name, ringo) G
  56. 56. Example G (R1, name, john) (R1, email, J@ed.ex) (R2, name, paul) (R1, webPage, ?W ) G (R2, name, paul) G µ∅ = { } (R3, name, ringo) G
  57. 57. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables:
  58. 58. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2).
  59. 59. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping.
  60. 60. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2
  61. 61. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2
  62. 62. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2 µ1 ∪ µ2 R1 john J@edu.ex
  63. 63. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2 µ1 ∪ µ2 R1 john J@edu.ex
  64. 64. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2 µ1 ∪ µ2 R1 john J@edu.ex µ1 ∪ µ3 R1 john P@edu.ex R2
  65. 65. Compatible mappings: mappings that can be merged. Definition The mappings µ1, µ2 are compatibles iff they agree in their shared variables: ◮ µ1(?X) = µ2(?X) for every ?X ∈ dom(µ1) ∩ dom(µ2). µ1 ∪ µ2 is also a mapping. Example ?X ?Y ?U ?V µ1 R1 john µ2 R1 J@edu.ex µ3 P@edu.ex R2 µ1 ∪ µ2 R1 john J@edu.ex µ1 ∪ µ3 R1 john P@edu.ex R2 µ∅ = { } is compatible with every mapping.
  66. 66. Sets of mappings and operations Let M1 and M2 be sets of mappings: Definition Join: M1 ⋊⋉ M2 ◮ {µ1 ∪ µ2 | µ1 ∈ M1, µ2 ∈ M2, and µ1, µ2 are compatibles} ◮ extending mappings in M1 with compatible mappings in M2 will be used to define AND
  67. 67. Sets of mappings and operations Let M1 and M2 be sets of mappings: Definition Join: M1 ⋊⋉ M2 ◮ {µ1 ∪ µ2 | µ1 ∈ M1, µ2 ∈ M2, and µ1, µ2 are compatibles} ◮ extending mappings in M1 with compatible mappings in M2 will be used to define AND Definition Union: M1 ∪ M2 ◮ {µ | µ ∈ M1 or µ ∈ M2} ◮ mappings in M1 plus mappings in M2 (the usual set union) will be used to define UNION
  68. 68. Sets of mappings and operations Definition Difference: M1 M2 ◮ {µ ∈ M1 | for all µ′ ∈ M2, µ and µ′ are not compatibles} ◮ mappings in M1 that cannot be extended with mappings in M2
  69. 69. Sets of mappings and operations Definition Difference: M1 M2 ◮ {µ ∈ M1 | for all µ′ ∈ M2, µ and µ′ are not compatibles} ◮ mappings in M1 that cannot be extended with mappings in M2 Definition Left outer join: M1 M2 = (M1 ⋊⋉ M2) ∪ (M1 M2) ◮ extension of mappings in M1 with compatible mappings in M2 ◮ plus the mappings in M1 that cannot be extended. will be used to define OPT
  70. 70. Semantics of general graph patterns Definition Given a graph G the evaluation of a pattern is recursively defined the base case is the evaluation of a triple pattern.
  71. 71. Semantics of general graph patterns Definition Given a graph G the evaluation of a pattern is recursively defined ◮ (P1 AND P2) G = P1 G ⋊⋉ P2 G the base case is the evaluation of a triple pattern.
  72. 72. Semantics of general graph patterns Definition Given a graph G the evaluation of a pattern is recursively defined ◮ (P1 AND P2) G = P1 G ⋊⋉ P2 G ◮ (P1 UNION P2) G = P1 G ∪ P2 G the base case is the evaluation of a triple pattern.
  73. 73. Semantics of general graph patterns Definition Given a graph G the evaluation of a pattern is recursively defined ◮ (P1 AND P2) G = P1 G ⋊⋉ P2 G ◮ (P1 UNION P2) G = P1 G ∪ P2 G ◮ (P1 OPT P2) G = P1 G P2 G the base case is the evaluation of a triple pattern.
  74. 74. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G
  75. 75. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G (?X, name, ?N) G ⋊⋉ (?X, email, ?E) G
  76. 76. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G (?X, name, ?N) G ⋊⋉ (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo
  77. 77. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G (?X, name, ?N) G ⋊⋉ (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex
  78. 78. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G (?X, name, ?N) G ⋊⋉ (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ⋊⋉ ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex
  79. 79. Example (AND) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) AND (?X, email, ?E)) G (?X, name, ?N) G ⋊⋉ (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ⋊⋉ ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex
  80. 80. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G
  81. 81. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G
  82. 82. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo
  83. 83. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex
  84. 84. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex
  85. 85. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul
  86. 86. Example (OPT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) OPT (?X, email, ?E)) G (?X, name, ?N) G (?X, email, ?E) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?X ?E µ4 R1 J@ed.ex µ5 R3 R@ed.ex ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul
  87. 87. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G
  88. 88. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G (?X, email, ?Info) G ∪ (?X, webPage, ?Info) G
  89. 89. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G (?X, email, ?Info) G ∪ (?X, webPage, ?Info) G ?X ?Info µ1 R1 J@ed.ex µ2 R3 R@ed.ex
  90. 90. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G (?X, email, ?Info) G ∪ (?X, webPage, ?Info) G ?X ?Info µ1 R1 J@ed.ex µ2 R3 R@ed.ex ?X ?Info µ3 R3 www.ringo.com
  91. 91. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G (?X, email, ?Info) G ∪ (?X, webPage, ?Info) G ?X ?Info µ1 R1 J@ed.ex µ2 R3 R@ed.ex ∪ ?X ?Info µ3 R3 www.ringo.com
  92. 92. Example (UNION) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, email, ?Info) UNION (?X, webPage, ?Info)) G (?X, email, ?Info) G ∪ (?X, webPage, ?Info) G ?X ?Info µ1 R1 J@ed.ex µ2 R3 R@ed.ex ∪ ?X ?Info µ3 R3 www.ringo.com ?X ?Info µ1 R1 J@ed.ex µ2 R3 R@ed.ex µ3 R3 www.ringo.com
  93. 93. Boolean filter expressions (value constraints) In filter expressions we consider ◮ the equality = among variables and RDF terms ◮ a unary predicate bound ◮ boolean combinations (∧, ∨, ¬) A mapping µ satisfies ◮ ?X = c if µ(?X) = c ◮ ?X =?Y if µ(?X) = µ(?Y ) ◮ bound(?X) if µ is defined in ?X, i.e. ?X ∈ dom(µ)
  94. 94. Satisfaction of value constraints ◮ If P is a graph pattern and R is a value constraint then (P FILTER R) is also a graph pattern.
  95. 95. Satisfaction of value constraints ◮ If P is a graph pattern and R is a value constraint then (P FILTER R) is also a graph pattern. Definition Given a graph G ◮ (P FILTER R) G = {µ ∈ P G | µ satisfies R} i.e. mappings in the evaluation of P that satisfy R.
  96. 96. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) FILTER (?N = ringo ∨ ?N = paul)) G
  97. 97. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) FILTER (?N = ringo ∨ ?N = paul)) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo
  98. 98. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) FILTER (?N = ringo ∨ ?N = paul)) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?N = ringo ∨ ?N = paul
  99. 99. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) ((?X, name, ?N) FILTER (?N = ringo ∨ ?N = paul)) G ?X ?N µ1 R1 john µ2 R2 paul µ3 R3 ringo ?N = ringo ∨ ?N = paul ?X ?N µ2 R2 paul µ3 R3 ringo
  100. 100. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (((?X, name, ?N) OPT (?X, email, ?E)) FILTER ¬ bound(?E)) G
  101. 101. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (((?X, name, ?N) OPT (?X, email, ?E)) FILTER ¬ bound(?E)) G ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul
  102. 102. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (((?X, name, ?N) OPT (?X, email, ?E)) FILTER ¬ bound(?E)) G ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul ¬ bound(?E)
  103. 103. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (((?X, name, ?N) OPT (?X, email, ?E)) FILTER ¬ bound(?E)) G ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul ¬ bound(?E) ?X ?N µ2 R2 paul
  104. 104. Example (FILTER) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (((?X, name, ?N) OPT (?X, email, ?E)) FILTER ¬ bound(?E)) G ?X ?N ?E µ1 ∪ µ4 R1 john J@ed.ex µ3 ∪ µ5 R3 ringo R@ed.ex µ2 R2 paul ¬ bound(?E) ?X ?N µ2 R2 paul (a non-monotonic query)
  105. 105. Why do we need/want to formalize SPARQL A formalization is beneficial ◮ clarifying corner cases ◮ helping in the implementation process ◮ providing solid foundations (we can actually prove properties!)
  106. 106. The evaluation decision problem Evaluation problem for SPARQL patterns Input: A mapping µ, an RDF graph G a graph pattern P Output: Is the mapping µ in the evaluation of pattern P over the RDF graph G
  107. 107. Brief complexity-theory reminder P
  108. 108. Brief complexity-theory reminder NP P
  109. 109. Brief complexity-theory reminder PSPACE NP P
  110. 110. Complexity of the evaluation problem Theorem (PAG09) For patterns using only the AND operator, the evaluation problem is in P
  111. 111. Complexity of the evaluation problem Theorem (PAG09) For patterns using only the AND operator, the evaluation problem is in P Theorem (SML10) For patterns using AND and UNION operators, the evaluation problem is NP-complete.
  112. 112. Complexity of the evaluation problem Theorem (PAG09) For patterns using only the AND operator, the evaluation problem is in P Theorem (SML10) For patterns using AND and UNION operators, the evaluation problem is NP-complete. Theorem (PAG09,SML10) For general patterns that include OPT operator, the evaluation problem is PSPACE-complete.
  113. 113. Complexity of the evaluation problem Theorem (PAG09) For patterns using only the AND operator, the evaluation problem is in P Theorem (SML10) For patterns using AND and UNION operators, the evaluation problem is NP-complete. Theorem (PAG09,SML10) For general patterns that include OPT operator, the evaluation problem is PSPACE-complete. Good news: evaluation in P if the query is fixed (data complexity)
  114. 114. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) if a variable occurs
  115. 115. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ if a variable occurs inside B
  116. 116. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT,
  117. 117. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ⇑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT, then the variable must also occur inside A.
  118. 118. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ⇑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT, then the variable must also occur inside A. Example (?Y , name, paul) OPT (?X, email, ?Z) AND (?X, name, john)
  119. 119. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ⇑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT, then the variable must also occur inside A. Example (?Y , name, paul) OPT (?X, email, ?Z) AND (?X, name, john) ↑
  120. 120. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ⇑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT, then the variable must also occur inside A. Example (?Y , name, paul) OPT (?X, email, ?Z) AND (?X, name, john) ↑ ↑
  121. 121. Well–designed patterns Can we find a natural fragment with better complexity? Definition An AND-OPT pattern is well–designed iff for every OPT in the pattern ( · · · · · · · · · · · · ( A OPT B ) · · · · · · · · · · · · ) ↑ ⇑ ↑ ↑ if a variable occurs inside B and anywhere outside the OPT, then the variable must also occur inside A. Example (?Y , name, paul) OPT (?X, email, ?Z) AND (?X, name, john) ↑ ↑
  122. 122. A bit more on complexity... PSPACE NP P
  123. 123. A bit more on complexity... PSPACE coNPNP P
  124. 124. Evaluation of well-designed patterns is in coNP-complete Theorem (PAG09) For AND-OPT well–designed graph patterns the evaluation problem is coNP-complete
  125. 125. Evaluation of well-designed patterns is in coNP-complete Theorem (PAG09) For AND-OPT well–designed graph patterns the evaluation problem is coNP-complete Well-designed patterns also allow to study static analysis: Theorem (LPPS12) Equivalence of well-designed SPARQL patterns is in NP
  126. 126. SELECT (a.k.a. projection) Besides graph patterns, SPARQL 1.0 allow result forms the most simple is SELECT Definition A SELECT query is an expression (SELECT W P) where P is a graph pattern and W is a set of variables, or *
  127. 127. SELECT (a.k.a. projection) Besides graph patterns, SPARQL 1.0 allow result forms the most simple is SELECT Definition A SELECT query is an expression (SELECT W P) where P is a graph pattern and W is a set of variables, or * The evaluation of a SELECT query against G is ◮ (SELECT W P) G = {µ|W | µ ∈ P G } where µ|W is the restriction of µ to domain W . ◮ (SELECT * P) G = P G
  128. 128. Example (SELECT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (SELECT {?N, ?E} ((?X, name, ?N) AND (?X, email, ?E))) G
  129. 129. Example (SELECT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (SELECT {?N, ?E} ((?X, name, ?N) AND (?X, email, ?E))) G SELECT{?N, ?E} ?X ?N ?E µ1 R1 john J@ed.ex µ2 R3 ringo R@ed.ex
  130. 130. Example (SELECT) G : (R1, name, john) (R2, name, paul) (R3, name, ringo) (R1, email, J@ed.ex) (R3, email, R@ed.ex) (R3, webPage, www.ringo.com) (SELECT {?N, ?E} ((?X, name, ?N) AND (?X, email, ?E))) G SELECT{?N, ?E} ?X ?N ?E µ1 R1 john J@ed.ex µ2 R3 ringo R@ed.ex ?N ?E µ1|{?N,?E} john J@ed.ex µ2|{?N,?E} ringo R@ed.ex
  131. 131. SPARQL 1.1 introduces several new features In SPARQL 1.1: ◮ (SELECT W P) can be used as any other graph pattern (ASK P) can be used as a constraint in FILTER ⇒ sub-queries ◮ Aggregations via ORDER-BY plus COUNT, SUM, etc. ◮ More important for us: Federation and Navigation
  132. 132. Outline Basics of SPARQL Syntax and Semantics of SPARQL 1.0 What is new in SPARQL 1.1 Federation: SERVICE operator Syntax and Semantics Evaluation of SERVICE queries Navigation: Property Paths Navigating graphs with regular expressions The history of paths (in SPARQL 1.1 specification) Evaluation procedures and complexity
  133. 133. SPARQL endpoints and the SERVICE operator ◮ SPARQL endpoints are services that accepts HTTP requests asking for SPARQL queries ◮ http://www.w3.org/wiki/SparqlEndpoints lists some ◮ SPARQL 1.1 allows to mix local and remote queries to endpoints via the SERVICE operator
  134. 134. SPARQL endpoints and the SERVICE operator ◮ SPARQL endpoints are services that accepts HTTP requests asking for SPARQL queries ◮ http://www.w3.org/wiki/SparqlEndpoints lists some ◮ SPARQL 1.1 allows to mix local and remote queries to endpoints via the SERVICE operator ?SELECT ?Author WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . }
  135. 135. SPARQL endpoints and the SERVICE operator ◮ SPARQL endpoints are services that accepts HTTP requests asking for SPARQL queries ◮ http://www.w3.org/wiki/SparqlEndpoints lists some ◮ SPARQL 1.1 allows to mix local and remote queries to endpoints via the SERVICE operator ?SELECT ?Author ?Place WHERE { ?Paper dc:creator ?Author . ?Paper dct:partOf ?Conf . ?Conf swrc:series conf:iswc . SERVICE dbpedia:service { ?Author dbpedia:birthPlace ?Place . } }
  136. 136. Syntax of SERVICE Syntax If c ∈ U, ?X is a variable, and P is a graph pattern then
  137. 137. Syntax of SERVICE Syntax If c ∈ U, ?X is a variable, and P is a graph pattern then ◮ (SERVICE c P) ◮ (SERVICE ?X P) are SERVICE graph patterns.
  138. 138. Syntax of SERVICE Syntax If c ∈ U, ?X is a variable, and P is a graph pattern then ◮ (SERVICE c P) ◮ (SERVICE ?X P) are SERVICE graph patterns. ◮ SERVICE graph pattern are included recursively in the algebra of graph patterns ◮ Variables are included in the SERVICE syntax to allow dynamic choosing of endpoints
  139. 139. Semantics of SERVICE We assume the existence of a partial function ep : U → RDF graphs intuitively, ep(c) is the (default) graph associated to the SPARQL endpoint defined by c
  140. 140. Semantics of SERVICE We assume the existence of a partial function ep : U → RDF graphs intuitively, ep(c) is the (default) graph associated to the SPARQL endpoint defined by c Definition Given an RDF graph G, an element c ∈ U, and a graph pattern P (SERVICE c P) G =
  141. 141. Semantics of SERVICE We assume the existence of a partial function ep : U → RDF graphs intuitively, ep(c) is the (default) graph associated to the SPARQL endpoint defined by c Definition Given an RDF graph G, an element c ∈ U, and a graph pattern P (SERVICE c P) G = P ep(c) if c ∈ dom(ep)
  142. 142. Semantics of SERVICE We assume the existence of a partial function ep : U → RDF graphs intuitively, ep(c) is the (default) graph associated to the SPARQL endpoint defined by c Definition Given an RDF graph G, an element c ∈ U, and a graph pattern P (SERVICE c P) G = P ep(c) if c ∈ dom(ep) {µ∅} otherwise
  143. 143. Semantics of SERVICE We assume the existence of a partial function ep : U → RDF graphs intuitively, ep(c) is the (default) graph associated to the SPARQL endpoint defined by c Definition Given an RDF graph G, an element c ∈ U, and a graph pattern P (SERVICE c P) G = P ep(c) if c ∈ dom(ep) {µ∅} otherwise but the interesting case is when the endpoint is a variable...
  144. 144. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G =
  145. 145. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G = P ep(c)
  146. 146. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G = P ep(c) ⋊⋉ {{?X → c}}
  147. 147. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G = c∈dom(ep) P ep(c) ⋊⋉ {{?X → c}}
  148. 148. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G = c∈dom(ep) P ep(c) ⋊⋉ {{?X → c}}
  149. 149. Semantics of SERVICE Definition Given an RDF graph G, a variable ?X, and a graph pattern P (SERVICE ?X P) G = c∈dom(ep) P ep(c) ⋊⋉ {{?X → c}} can we effectively evaluate a SERVICE query?
  150. 150. SERVICE example Some queries/patterns can be safely evaluated:
  151. 151. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)))
  152. 152. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))) ◮ ((SERVICE ?Y (?N, email, ?E)) AND (?X, service address, ?Y ))
  153. 153. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))) ◮ ((SERVICE ?Y (?N, email, ?E)) AND (?X, service address, ?Y )) In both cases, the SERVICE variable is controlled by the data in the initial graph
  154. 154. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))) ◮ ((SERVICE ?Y (?N, email, ?E)) AND (?X, service address, ?Y )) In both cases, the SERVICE variable is controlled by the data in the initial graph There is natural way of evaluating the query:
  155. 155. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))) ◮ ((SERVICE ?Y (?N, email, ?E)) AND (?X, service address, ?Y )) In both cases, the SERVICE variable is controlled by the data in the initial graph There is natural way of evaluating the query: ◮ Evaluate first (?X, service address, ?Y )
  156. 156. SERVICE example Some queries/patterns can be safely evaluated: ◮ ((?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))) ◮ ((SERVICE ?Y (?N, email, ?E)) AND (?X, service address, ?Y )) In both cases, the SERVICE variable is controlled by the data in the initial graph There is natural way of evaluating the query: ◮ Evaluate first (?X, service address, ?Y ) ◮ only for the obtained mappings evaluate the SERVICE on endpoint µ(?Y )
  157. 157. Unbounded SERVICE queries What about this pattern? (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))
  158. 158. Unbounded SERVICE queries What about this pattern? (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) :-(
  159. 159. Unbounded SERVICE queries What about this pattern? (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) :-( Idea: force the SERVICE variable to be bound in every solution.
  160. 160. Boundedness Definition A variable ?X is bound in graph pattern P if for every graph G and every µ ∈ P G it holds that: ◮ ?X ∈ dom(µ), and ◮ µ(?X) is a value in G.
  161. 161. Boundedness Definition A variable ?X is bound in graph pattern P if for every graph G and every µ ∈ P G it holds that: ◮ ?X ∈ dom(µ), and ◮ µ(?X) is a value in G. We only need a procedure to ensure that every variable mentioned in SERVICE is bounded!
  162. 162. Boundedness Definition A variable ?X is bound in graph pattern P if for every graph G and every µ ∈ P G it holds that: ◮ ?X ∈ dom(µ), and ◮ µ(?X) is a value in G. We only need a procedure to ensure that every variable mentioned in SERVICE is bounded! Oh wait...
  163. 163. Boundedness Definition A variable ?X is bound in graph pattern P if for every graph G and every µ ∈ P G it holds that: ◮ ?X ∈ dom(µ), and ◮ µ(?X) is a value in G. We only need a procedure to ensure that every variable mentioned in SERVICE is bounded! Oh wait... Theorem (BAC11) The problem of checking if a variable is bound in a graph pattern is undecidable.
  164. 164. Boundedness Definition A variable ?X is bound in graph pattern P if for every graph G and every µ ∈ P G it holds that: ◮ ?X ∈ dom(µ), and ◮ µ(?X) is a value in G. We only need a procedure to ensure that every variable mentioned in SERVICE is bounded! Oh wait... Theorem (BAC11) The problem of checking if a variable is bound in a graph pattern is undecidable. :-(
  165. 165. Undecidability of boundedness Proof idea ◮ From [AG08]: it is undecidable to check if a SPARQL pattern P is satisfiable (if P G = ∅ for some G).
  166. 166. Undecidability of boundedness Proof idea ◮ From [AG08]: it is undecidable to check if a SPARQL pattern P is satisfiable (if P G = ∅ for some G). ◮ Assume P does not mention ?X, and let Q = ((?X, ?Y , ?Z) UNION P):
  167. 167. Undecidability of boundedness Proof idea ◮ From [AG08]: it is undecidable to check if a SPARQL pattern P is satisfiable (if P G = ∅ for some G). ◮ Assume P does not mention ?X, and let Q = ((?X, ?Y , ?Z) UNION P): ?X is bound in Q iff P is not satisfiable.
  168. 168. Undecidability of boundedness Proof idea ◮ From [AG08]: it is undecidable to check if a SPARQL pattern P is satisfiable (if P G = ∅ for some G). ◮ Assume P does not mention ?X, and let Q = ((?X, ?Y , ?Z) UNION P): ?X is bound in Q iff P is not satisfiable. Undecidable: the SPARQL engine cannot check for boundedness...
  169. 169. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then
  170. 170. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2)
  171. 171. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then
  172. 172. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1)
  173. 173. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then
  174. 174. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2)
  175. 175. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then
  176. 176. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then SB(P) = W ∩ SB(P1)
  177. 177. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then SB(P) = W ∩ SB(P1) ◮ if P is a SERVICE pattern, then
  178. 178. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then SB(P) = W ∩ SB(P1) ◮ if P is a SERVICE pattern, then SB(P) = ∅
  179. 179. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then SB(P) = W ∩ SB(P1) ◮ if P is a SERVICE pattern, then SB(P) = ∅ Proposition (BAC11) If ?X ∈ SB(P) then ?X is bound in P.
  180. 180. A sufficient condition for boundedness Definition [BAC11] The set of strongly bounded variables in a pattern P, denoted by SB(P) is defined recursively as follows. ◮ if P is a triple pattern t, then SB(P) = var(t) ◮ if P = (P1 AND P2), then SB(P) = SB(P1) ∪ SB(P2) ◮ if P = (P1 OPT P2), then SB(P) = SB(P1) ◮ if P = (P1 UNION P2), then SB(P) = SB(P1) ∩ SB(P2) ◮ if P = (SELECT W P1), then SB(P) = W ∩ SB(P1) ◮ if P is a SERVICE pattern, then SB(P) = ∅ Proposition (BAC11) If ?X ∈ SB(P) then ?X is bound in P. (SB(P) can be efficiently computed)
  181. 181. (Strongly) boundedness is not enough Are we happy now? P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) .
  182. 182. (Strongly) boundedness is not enough Are we happy now? P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) . ◮ ?Y is not bound in P1 (nor strongly bound)
  183. 183. (Strongly) boundedness is not enough Are we happy now? P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) . ◮ ?Y is not bound in P1 (nor strongly bound) ◮ nevertheless there is a natural evaluation of this pattern
  184. 184. (Strongly) boundedness is not enough Are we happy now? P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) . ◮ ?Y is not bound in P1 (nor strongly bound) ◮ nevertheless there is a natural evaluation of this pattern (?Y is bounded in the important part!)
  185. 185. (Strongly) boundedness is not enough Are we happy now?
  186. 186. (Strongly) boundedness is not enough Are we happy now? P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F)) .
  187. 187. (Strongly) boundedness is not enough Are we happy now? P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F)) . ◮ ?U1 and ?U2 are strongly bounded, but
  188. 188. (Strongly) boundedness is not enough Are we happy now? P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F)) . ◮ ?U1 and ?U2 are strongly bounded, but ◮ can we effectively evaluate this query? (?U2 is unbounded in the important part!)
  189. 189. (Strongly) boundedness is not enough Are we happy now? P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F)) . ◮ ?U1 and ?U2 are strongly bounded, but ◮ can we effectively evaluate this query? (?U2 is unbounded in the important part!) we need to define what the important part is...
  190. 190. Parse tree of a pattern We need first to formalize the tree of subexpressions of a pattern (?Y , a, ?Z) UNION (?X, b, c) AND (SERVICE ?X (?Y , a, ?Z))
  191. 191. Parse tree of a pattern We need first to formalize the tree of subexpressions of a pattern (?Y , a, ?Z) UNION (?X, b, c) AND (SERVICE ?X (?Y , a, ?Z)) u6 : (?Y , a, ?Z) u1 : ((?Y , a, ?Z) UNION ((?X, b, c) AND (SERVICE ?X (?Y , a, ?Z)))) u2 : (?Y , a, ?Z) u3 : ((?X, b, c) AND (SERVICE ?X (?Y , a, ?Z))) u4 : (?X, b, c) u5 : (SERVICE ?X (?Y , a, ?Z))
  192. 192. Parse tree of a pattern We need first to formalize the tree of subexpressions of a pattern (?Y , a, ?Z) UNION (?X, b, c) AND (SERVICE ?X (?Y , a, ?Z)) u6 : (?Y , a, ?Z) u1 : ((?Y , a, ?Z) UNION ((?X, b, c) AND (SERVICE ?X (?Y , a, ?Z)))) u2 : (?Y , a, ?Z) u3 : ((?X, b, c) AND (SERVICE ?X (?Y , a, ?Z))) u4 : (?X, b, c) u5 : (SERVICE ?X (?Y , a, ?Z)) We denote by T (P) the tree of subexpressions of P.
  193. 193. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that:
  194. 194. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X is bound in P2, and
  195. 195. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X is bound in P2, and 2. P1 is service-bound.
  196. 196. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X is bound in P2, and 2. P1 is service-bound. Unfortunately... (and not so surprisingly anymore)
  197. 197. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X is bound in P2, and 2. P1 is service-bound. Unfortunately... (and not so surprisingly anymore) Theorem (BAC11) Checking if a pattern is service-bound is undecidable.
  198. 198. Service-boundedness Notion of boundedness considering the important part Definition A pattern P is service-bound if for every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X is bound in P2, and 2. P1 is service-bound. Unfortunately... (and not so surprisingly anymore) Theorem (BAC11) Checking if a pattern is service-bound is undecidable. Exercise: prove the theorem.
  199. 199. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion.
  200. 200. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion. Definition A pattern P is service-safe if of every node u in T (P) with label (SERVICE ?X P1) it holds that:
  201. 201. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion. Definition A pattern P is service-safe if of every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X ∈ SB(P2), and
  202. 202. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion. Definition A pattern P is service-safe if of every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X ∈ SB(P2), and 2. P1 is service-safe.
  203. 203. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion. Definition A pattern P is service-safe if of every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X ∈ SB(P2), and 2. P1 is service-safe. Proposition (BAC11) If P is service-safe the it is service-bound.
  204. 204. Service-safeness We need a decidable sufficient condition. Idea: replace bound by SB(·) in the previous notion. Definition A pattern P is service-safe if of every node u in T (P) with label (SERVICE ?X P1) it holds that: 1. There exists an ancestor of u in T (P) with label P2 s.t. ?X ∈ SB(P2), and 2. P1 is service-safe. Proposition (BAC11) If P is service-safe the it is service-bound. We finally have our desired decidable condition.
  205. 205. Applying service-safeness P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E))
  206. 206. Applying service-safeness P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) is service-safe.
  207. 207. Applying service-safeness P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) is service-safe. P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F))
  208. 208. Applying service-safeness P1 = (?X, service description, ?Z) UNION (?X, service address, ?Y ) AND (SERVICE ?Y (?N, email, ?E)) is service-safe. P2 = (?U1, related with, ?U2) AND SERVICE ?U1 (?N, email, ?E) OPT (SERVICE ?U2 (?N, phone, ?F)) is not service-safe.
  209. 209. Closing words on SERVICE SERVICE: several interesting research question ◮ Optimization (rewriting, reordering, containment?) ◮ Cost analysis: in practice we cannot query all that we may want ◮ Different endpoints provide different completeness certificates (regarding the data they return) ◮ Several implementations challenges
  210. 210. Outline Basics of SPARQL Syntax and Semantics of SPARQL 1.0 What is new in SPARQL 1.1 Federation: SERVICE operator Syntax and Semantics Evaluation of SERVICE queries Navigation: Property Paths Navigating graphs with regular expressions The history of paths (in SPARQL 1.1 specification) Evaluation procedures and complexity
  211. 211. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2)
  212. 212. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections?
  213. 213. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic
  214. 214. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic You (should) already know that Datalog can express this query.
  215. 215. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic You (should) already know that Datalog can express this query. We can consider a new predicate reach and the program
  216. 216. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic You (should) already know that Datalog can express this query. We can consider a new predicate reach and the program (?X, reach, ?Y ) ← (?X, connected, ?Y )
  217. 217. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic You (should) already know that Datalog can express this query. We can consider a new predicate reach and the program (?X, reach, ?Y ) ← (?X, connected, ?Y ) (?X, reach, ?Y ) ← (?X, reach, ?Z), (?Z, connected, ?Y )
  218. 218. SPARQL 1.0 has very limited navigational capabilities Assume a graph with cities and connections with RDF triples like: (C1, connected, C2) query: is city B reachable from A by a sequence of connections? ◮ Known fact: SPARQL 1.0 cannot express this query! ◮ Follows easily from locality of FO-logic You (should) already know that Datalog can express this query. We can consider a new predicate reach and the program (?X, reach, ?Y ) ← (?X, connected, ?Y ) (?X, reach, ?Y ) ← (?X, reach, ?Z), (?Z, connected, ?Y ) but do we want to integrate Datalog and SPARQL?
  219. 219. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf
  220. 220. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X :friendOf ?Y . ?Y :name "Maria" . }
  221. 221. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X :friendOf ?Y . ?Y :name "Maria" . }
  222. 222. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X (:friendOf)* ?Y . ?Y :name "Maria" . }
  223. 223. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X (:friendOf)* ?Y . ?Y :name "Maria" . }
  224. 224. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X (:friendOf)* ?Y . ← SPARQL 1.1 property path ?Y :name "Maria" . }
  225. 225. SPARQL 1.1 provides an alternative way for navigating URI2 :email seba@puc.cl :name :email :phone :name :friendOf juan@utexas.edu URI1 Seba 446928888 Juan Claudio :name :name Maria URI0 :friendOf URI3 :friendOf SELECT ?X WHERE { ?X (:friendOf)* ?Y . ← SPARQL 1.1 property path ?Y :name "Maria" . } Idea: navigate RDF graphs using regular expressions
  226. 226. General navigation using regular expressions Regular expressions define sets of strings using ◮ concatenation: / ◮ disjunction: | ◮ Kleene star: * Example Consider strings composed of symbols a and b a/(b)*/a defines strings of the form abbb · · · bbba.
  227. 227. General navigation using regular expressions Regular expressions define sets of strings using ◮ concatenation: / ◮ disjunction: | ◮ Kleene star: * Example Consider strings composed of symbols a and b a/(b)*/a defines strings of the form abbb · · · bbba. Idea: use regular expressions to define paths ◮ a path p satisfies a regular expression r if the string composed of the sequence of edges of p satisfies expression r
  228. 228. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John
  229. 229. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X }
  230. 230. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection
  231. 231. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X Paris }
  232. 232. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)* Paris }
  233. 233. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)*/:bus Paris }
  234. 234. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)*/:bus/(:train|:plane)* Paris }
  235. 235. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)*/:bus/(:train|:plane)* Paris } Exercise: cities that reach Paris with an even number of connections
  236. 236. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)*/:bus/(:train|:plane)* Paris } Exercise: cities that reach Paris with an even number of connections Mixing regular expressions and SPARQL operators gives interesting expressive power: Persons in my professional network that attended the same school
  237. 237. Interesting navigational queries ◮ RDF graph with :father, :mother edges: ancestors of John { John (:father|:mother)* ?X } ◮ Connections between cities via :train, :bus, :plane Cities that reach Paris with exactly one :bus connection { ?X (:train|:plane)*/:bus/(:train|:plane)* Paris } Exercise: cities that reach Paris with an even number of connections Mixing regular expressions and SPARQL operators gives interesting expressive power: Persons in my professional network that attended the same school { ?X (:conn)* ?Y . ?X (:conn)* ?Z . ?Y :sameSchool ?Z }
  238. 238. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries
  239. 239. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths
  240. 240. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths ◮ Late 2010: several discussions on possible drawbacks on the non-standard definition by the W3C
  241. 241. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths ◮ Late 2010: several discussions on possible drawbacks on the non-standard definition by the W3C ◮ Early 2011: first formal semantics by the W3C
  242. 242. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths ◮ Late 2010: several discussions on possible drawbacks on the non-standard definition by the W3C ◮ Early 2011: first formal semantics by the W3C ◮ Late 2011: empirical and theoretical study on SPARQL 1.1 property paths showing unfeasibility of evaluation
  243. 243. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths ◮ Late 2010: several discussions on possible drawbacks on the non-standard definition by the W3C ◮ Early 2011: first formal semantics by the W3C ◮ Late 2011: empirical and theoretical study on SPARQL 1.1 property paths showing unfeasibility of evaluation ◮ Mid 2012: semantics change to overcome the raised issues
  244. 244. As always, we need a (formal) semantics ◮ Regular expressions in SPARQL queries seem reasonable ◮ We need to agree in the meaning of these new queries A bit of history on W3C standardization of property paths: ◮ Mid 2010: W3C defines an informal semantics for paths ◮ Late 2010: several discussions on possible drawbacks on the non-standard definition by the W3C ◮ Early 2011: first formal semantics by the W3C ◮ Late 2011: empirical and theoretical study on SPARQL 1.1 property paths showing unfeasibility of evaluation ◮ Mid 2012: semantics change to overcome the raised issues The following experimental study is based on [ACP12].
  245. 245. SPARQL 1.1 implementations had a poor performance Data: ◮ cliques (complete graphs) of different size ◮ from 2 nodes (87 bytes) to 13 nodes (970 bytes) :p :a0 :a1 :a3 :p :p :p :p :p :a2 RDF clique with 4 nodes (127 bytes)
  246. 246. SPARQL 1.1 implementations had a poor performance 1 10 100 1000 2 4 6 8 10 12 14 16 ARQ + + + + + + + + + + + + RDFQ × × × × × × × × × × KGram ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ ∗ Sesame [ACP12] SELECT * WHERE { :a0 (:p)* :a1 }
  247. 247. Poor performance with real Web data of small size Data: ◮ Social Network data given by foaf:knows links ◮ Crawled from Axel Polleres’ foaf document (3 steps) ◮ Different documents, deleting some nodes foaf:knows axel:me ivan:me bizer:chris richard:cygri ... · · · andreas:ah · · · · · ·
  248. 248. Poor performance with real Web data of small size SELECT * WHERE { axel:me (foaf:knows)* ?x }
  249. 249. Poor performance with real Web data of small size SELECT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame 9.2KB 5.13 75.70 313.37 – 10.9KB 8.20 325.83 – – 11.4KB 65.87 – – – 13.2KB 292.43 – – – 14.8KB – – – – 17.2KB – – – – 20.5KB – – – – 25.8KB – – – – [ACP12] (time in seconds, timeout = 1hr)
  250. 250. Poor performance with real Web data of small size SELECT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame 9.2KB 5.13 75.70 313.37 – 10.9KB 8.20 325.83 – – 11.4KB 65.87 – – – 13.2KB 292.43 – – – 14.8KB – – – – 17.2KB – – – – 20.5KB – – – – 25.8KB – – – – [ACP12] (time in seconds, timeout = 1hr) Is this a problem of these particular implementations?
  251. 251. This is a problem of the specification [ACP12] Any implementation that follows SPARQL 1.1 standard (as of January 2012) is doomed to show the same behavior
  252. 252. This is a problem of the specification [ACP12] Any implementation that follows SPARQL 1.1 standard (as of January 2012) is doomed to show the same behavior The main sources of complexity is counting
  253. 253. This is a problem of the specification [ACP12] Any implementation that follows SPARQL 1.1 standard (as of January 2012) is doomed to show the same behavior The main sources of complexity is counting Impact on W3C standard: ◮ Standard semantics of SPARQL 1.1 property paths was changed in July 2012 to overcome these issues
  254. 254. SPARQL 1.1 property paths match regular expressions but also count :a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X }
  255. 255. SPARQL 1.1 property paths match regular expressions but also count :a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X } ?X :a :b :c :d :d
  256. 256. SPARQL 1.1 property paths match regular expressions but also count :a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X } ?X :a :b :c :d :d
  257. 257. SPARQL 1.1 property paths match regular expressions but also count :p:a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X } ?X :a :b :c :d :d
  258. 258. SPARQL 1.1 property paths match regular expressions but also count :p:a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X } ?X :a :b :c :d :d :c :d
  259. 259. SPARQL 1.1 property paths match regular expressions but also count :p:a :b :c :p :p :p :p :d SELECT ?X WHERE { :a (:p)* ?X } ?X :a :b :c :d :d :c :d But what if we have cycles?
  260. 260. SPARQL 1.1 document provides a special procedure to handle cycles (and make the count) Evaluation of path* “the algorithm extends the multiset of results by one application of path. If a node has been visited for path, it is not a candidate for another step. A node can be visited multiple times if different paths visit it.” SPARQL 1.1 Last Call (Jan 2012)
  261. 261. SPARQL 1.1 document provides a special procedure to handle cycles (and make the count) Evaluation of path* “the algorithm extends the multiset of results by one application of path. If a node has been visited for path, it is not a candidate for another step. A node can be visited multiple times if different paths visit it.” SPARQL 1.1 Last Call (Jan 2012) ◮ W3C document provides a procedure (ArbitraryLengthPath) ◮ This procedure was formalized in [ACP12]
  262. 262. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  263. 263. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } n # Sol. 9 13,700 10 109,601 11 986,410 12 9,864,101 13 – every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  264. 264. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } n # Sol. 9 13,700 10 109,601 11 986,410 12 9,864,101 13 – every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  265. 265. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } n # Sol. 9 13,700 10 109,601 11 986,410 12 9,864,101 13 – n # Sol 2 1 3 6 4 305 5 418,576 6 – every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  266. 266. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } { :a0 (((:p)*)*)* :a1 } n # Sol. 9 13,700 10 109,601 11 986,410 12 9,864,101 13 – n # Sol 2 1 3 6 4 305 5 418,576 6 – every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  267. 267. Counting the number of solutions... Data: Clique of size n { :a0 (:p)* :a1 } { :a0 ((:p)*)* :a1 } { :a0 (((:p)*)*)* :a1 } n # Sol. 9 13,700 10 109,601 11 986,410 12 9,864,101 13 – n # Sol 2 1 3 6 4 305 5 418,576 6 – n # Sol. 2 1 3 42 4 – every solution is a copy of the empty mapping µ∅ (| | in ARQ)
  268. 268. More on counting the number of solutions... Data: foaf links crawled from the Web { axel:me (foaf:knows)* ?x }
  269. 269. More on counting the number of solutions... Data: foaf links crawled from the Web { axel:me (foaf:knows)* ?x } File # URIs # Sol. Output Size 9.2KB 38 29,817 2MB 10.9KB 43 122,631 8.4MB 11.4KB 47 1,739,331 120MB 13.2KB 52 8,511,943 587MB 14.8KB 54 – –
  270. 270. More on counting the number of solutions... Data: foaf links crawled from the Web { axel:me (foaf:knows)* ?x } File # URIs # Sol. Output Size 9.2KB 38 29,817 2MB 10.9KB 43 122,631 8.4MB 11.4KB 47 1,739,331 120MB 13.2KB 52 8,511,943 587MB 14.8KB 54 – – What is really happening here?
  271. 271. More on counting the number of solutions... Data: foaf links crawled from the Web { axel:me (foaf:knows)* ?x } File # URIs # Sol. Output Size 9.2KB 38 29,817 2MB 10.9KB 43 122,631 8.4MB 11.4KB 47 1,739,331 120MB 13.2KB 52 8,511,943 587MB 14.8KB 54 – – What is really happening here? Theory can help!
  272. 272. A bit more on complexity classes... Complexity can be measured by using counting-complexity classes NP #P Sat: is a propositional CountSat: how many assignments formula satisfiable? satisfy a propositional formula?
  273. 273. A bit more on complexity classes... Complexity can be measured by using counting-complexity classes NP #P Sat: is a propositional CountSat: how many assignments formula satisfiable? satisfy a propositional formula? Formally A function f (·) on strings is in #P if there exists a polynomial-time non-deterministic TM M such that f (x) = number of accepting computations of M with input x
  274. 274. A bit more on complexity classes... Complexity can be measured by using counting-complexity classes NP #P Sat: is a propositional CountSat: how many assignments formula satisfiable? satisfy a propositional formula? Formally A function f (·) on strings is in #P if there exists a polynomial-time non-deterministic TM M such that f (x) = number of accepting computations of M with input x ◮ CountSat is #P-complete
  275. 275. Counting problem for property paths CountW3C Input: RDF graph G Property path triple { :a path :b } Output: Count the number of solutions of { :a path :b } over G (according to the semantics proposed by W3C)
  276. 276. The complexity of property paths is intractable Theorem (ACP12) CountW3C is outside #P
  277. 277. The complexity of property paths is intractable Theorem (ACP12) CountW3C is outside #P CountW3C is hard to solve even if P = NP
  278. 278. A doubly exponential lower bound for counting ◮ Let paths be a property path of the form (· · · ((:p)*)*)· · ·)* with s nested stars
  279. 279. A doubly exponential lower bound for counting ◮ Let paths be a property path of the form (· · · ((:p)*)*)· · ·)* with s nested stars ◮ Let Kn be a clique with n nodes
  280. 280. A doubly exponential lower bound for counting ◮ Let paths be a property path of the form (· · · ((:p)*)*)· · ·)* with s nested stars ◮ Let Kn be a clique with n nodes ◮ Let CountClique(s, n) be the number of solutions of { :a0 paths :a1 } over Kn
  281. 281. A doubly exponential lower bound for counting ◮ Let paths be a property path of the form (· · · ((:p)*)*)· · ·)* with s nested stars ◮ Let Kn be a clique with n nodes ◮ Let CountClique(s, n) be the number of solutions of { :a0 paths :a1 } over Kn Lemma (ACP12) CountClique(s, n) ≥ (n − 2)!(n−1)s−1
  282. 282. A doubly exponential lower bound for counting ◮ Let paths be a property path of the form (· · · ((:p)*)*)· · ·)* with s nested stars ◮ Let Kn be a clique with n nodes ◮ Let CountClique(s, n) be the number of solutions of { :a0 paths :a1 } over Kn Lemma (ACP12) CountClique(s, n) ≥ (n − 2)!(n−1)s−1 In [ACP12]: A recursive formula for calculating CountClique(s, n)
  283. 283. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – 7 – 8 –
  284. 284. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – 7 – 8 –
  285. 285. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – 7 – 8 –
  286. 286. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – 7 – 8 –
  287. 287. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – 7 – 8 –
  288. 288. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – ← 28 × 109 7 – 8 –
  289. 289. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – ← 28 × 109 7 – ← 144 × 1015 8 –
  290. 290. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – ← 28 × 109 7 – ← 144 × 1015 8 – ← 79 × 1024
  291. 291. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – ← 28 × 109 7 – ← 144 × 1015 8 – ← 79 × 1024 79 Yottabytes for the answer over a file of 379 bytes
  292. 292. We can explain the experimental results CountClique(s, n) allows to fill in the blanks { :a0 ((:p)*)* :a1 } n # Sol. 2 1 3 6 4 305 5 418,576 6 – ← 28 × 109 7 – ← 144 × 1015 8 – ← 79 × 1024 79 Yottabytes for the answer over a file of 379 bytes 1 Yottabyte > the estimated capacity of all digital storage in the world
  293. 293. Data complexity of property path is still intractable Common assumption in Databases: ◮ queries are much smaller than data sources
  294. 294. Data complexity of property path is still intractable Common assumption in Databases: ◮ queries are much smaller than data sources Data complexity ◮ measure the complexity considering the query fixed
  295. 295. Data complexity of property path is still intractable Common assumption in Databases: ◮ queries are much smaller than data sources Data complexity ◮ measure the complexity considering the query fixed ◮ Data complexity of SQL, XPath, SPARQL 1.0, etc. are all polynomial
  296. 296. Data complexity of property path is still intractable Common assumption in Databases: ◮ queries are much smaller than data sources Data complexity ◮ measure the complexity considering the query fixed ◮ Data complexity of SQL, XPath, SPARQL 1.0, etc. are all polynomial Theorem (ACP12) Data complexity of CountW3C is #P-complete
  297. 297. Data complexity of property path is still intractable Common assumption in Databases: ◮ queries are much smaller than data sources Data complexity ◮ measure the complexity considering the query fixed ◮ Data complexity of SQL, XPath, SPARQL 1.0, etc. are all polynomial Theorem (ACP12) Data complexity of CountW3C is #P-complete Corollary SPARQL 1.1 query evaluation (as of January 2012) is intractable in Data Complexity
  298. 298. Existential semantics: a possible alternative Possible solution Do not count Just check whether there exists a path satisfying the property path expression
  299. 299. Existential semantics: a possible alternative Possible solution Do not count Just check whether there exists a path satisfying the property path expression Years of experiences (theory and practice) in: ◮ Graph Databases ◮ XML ◮ SPARQL 1.0 (PSPARQL, Gleen) + equivalent regular expressions giving equivalent results
  300. 300. Existential semantics: decision problems Input: RDF graph G Property path triple { :a path :b } ExistsPath Question: Is there a path from :a to :b in G satisfying the regular expression path? ExistsW3C Question: Is the number of solutions of { :a path :b } over G greater than 0 (according to W3C semantics)?
  301. 301. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory:
  302. 302. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state
  303. 303. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state 2. construct from path an NFA Apath
  304. 304. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state 2. construct from path an NFA Apath 3. construct the product automaton G × Apath:
  305. 305. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state 2. construct from path an NFA Apath 3. construct the product automaton G × Apath: ◮ whenever (x, r, y) ∈ G and (p, r, q) is a transition in Apath add a transition ((x, p), r, (y, q)) to G × Apath
  306. 306. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state 2. construct from path an NFA Apath 3. construct the product automaton G × Apath: ◮ whenever (x, r, y) ∈ G and (p, r, q) is a transition in Apath add a transition ((x, p), r, (y, q)) to G × Apath 4. check if we can go from (:a, q0) to (:b, qf ) in G × Apath
  307. 307. Evaluating existential paths is tractable Theorem (well-known result) ExistsPath can be solved in O(|G| × |path|) Can be proved by using automata theory: 1. consider G as an NFA with :a initial state and :b final state 2. construct from path an NFA Apath 3. construct the product automaton G × Apath: ◮ whenever (x, r, y) ∈ G and (p, r, q) is a transition in Apath add a transition ((x, p), r, (y, q)) to G × Apath 4. check if we can go from (:a, q0) to (:b, qf ) in G × Apath with q0 initial state of Apath and qf some final state of Apath
  308. 308. Relationship between ExistsPath and ExistsW3C Theorem (ACP12) ExistsPath and ExistsW3C are equivalent decision problems
  309. 309. Relationship between ExistsPath and ExistsW3C Theorem (ACP12) ExistsPath and ExistsW3C are equivalent decision problems Corollary (ACP12) ExistsW3C can be solved in O(|G| × |path|)
  310. 310. So there are possibilities for optimization SPARQL includes an operator to eliminate duplicates (DISTINCT)
  311. 311. So there are possibilities for optimization SPARQL includes an operator to eliminate duplicates (DISTINCT) Corollary Property path queries with SELECT DISTINCT can be efficiently evaluated
  312. 312. So there are possibilities for optimization SPARQL includes an operator to eliminate duplicates (DISTINCT) Corollary Property path queries with SELECT DISTINCT can be efficiently evaluated And we can also use DISTINCT over general queries Theorem SELECT DISTINCT SPARQL 1.1 queries are tractable in Data Complexity
  313. 313. SPARQL 1.1 implementations do not take advantage of SELECT DISTINCT SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame Psparql Gleen
  314. 314. SPARQL 1.1 implementations do not take advantage of SELECT DISTINCT SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame Psparql Gleen
  315. 315. SPARQL 1.1 implementations do not take advantage of SELECT DISTINCT SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame Psparql Gleen 9.2KB 2.24 47.31 2.37 – 0.29 1.39 10.9KB 2.60 204.95 6.43 – 0.30 1.32 11.4KB 6.88 3222.47 80.73 – 0.30 1.34 13.2KB 24.42 – 394.61 – 0.31 1.38 14.8KB – – – – 0.33 1.38 17.2KB – – – – 0.35 1.42 20.5KB – – – – 0.44 1.50 25.8KB – – – – 0.45 1.52
  316. 316. SPARQL 1.1 implementations do not take advantage of SELECT DISTINCT SELECT DISTINCT * WHERE { axel:me (foaf:knows)* ?x } Input ARQ RDFQ Kgram Sesame Psparql Gleen 9.2KB 2.24 47.31 2.37 – 0.29 1.39 10.9KB 2.60 204.95 6.43 – 0.30 1.32 11.4KB 6.88 3222.47 80.73 – 0.30 1.34 13.2KB 24.42 – 394.61 – 0.31 1.38 14.8KB – – – – 0.33 1.38 17.2KB – – – – 0.35 1.42 20.5KB – – – – 0.44 1.50 25.8KB – – – – 0.45 1.52 Optimization possibilities can remain hidden in a complicated specification
  317. 317. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression)
  318. 318. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression) For example, consider path1, path2, and path3 not using * then when evaluating { :a path1/(path2)*/path3 :b } one should (intuitively):
  319. 319. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression) For example, consider path1, path2, and path3 not using * then when evaluating { :a path1/(path2)*/path3 :b } one should (intuitively): ◮ consider all the paths from :a to some intermediate :c1 satisfying path1
  320. 320. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression) For example, consider path1, path2, and path3 not using * then when evaluating { :a path1/(path2)*/path3 :b } one should (intuitively): ◮ consider all the paths from :a to some intermediate :c1 satisfying path1 ◮ check if there exists :c2 reachable from :c1 following (path2)*
  321. 321. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression) For example, consider path1, path2, and path3 not using * then when evaluating { :a path1/(path2)*/path3 :b } one should (intuitively): ◮ consider all the paths from :a to some intermediate :c1 satisfying path1 ◮ check if there exists :c2 reachable from :c1 following (path2)* ◮ consider all the paths from :c2 to :b satisfying path3
  322. 322. New semantics for property paths (July 2012) ◮ Paths constructed only from / and | should be counted ◮ As soon as * is used, duplicates are eliminated (from that part of the expression) For example, consider path1, path2, and path3 not using * then when evaluating { :a path1/(path2)*/path3 :b } one should (intuitively): ◮ consider all the paths from :a to some intermediate :c1 satisfying path1 ◮ check if there exists :c2 reachable from :c1 following (path2)* ◮ consider all the paths from :c2 to :b satisfying path3 ◮ make the count (to produce copies)
  323. 323. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives?
  324. 324. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide.
  325. 325. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles
  326. 326. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles ◮ count the number of simple paths
  327. 327. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles ◮ count the number of simple paths ◮ does someone from the audience have another in mind?
  328. 328. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles ◮ count the number of simple paths ◮ does someone from the audience have another in mind? In all the above cases you have to decide what exactly you count:
  329. 329. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles ◮ count the number of simple paths ◮ does someone from the audience have another in mind? In all the above cases you have to decide what exactly you count: ◮ the number of paths satisfying the expression? ◮ the number of ways that the expr can be satisfied? (W3C)
  330. 330. Is the new semantics the right one? W3C definitely wants to count paths. Are there more reasonable alternatives? ◮ to have different operators for counting (e.g. . and ||) so the user can decide. ◮ the honest approach: just make the count and output infininty in the presence of cycles ◮ count the number of simple paths ◮ does someone from the audience have another in mind? In all the above cases you have to decide what exactly you count: ◮ the number of paths satisfying the expression? ◮ the number of ways that the expr can be satisfied? (W3C) (they are not the same! consider for example (a|a) )
  331. 331. Several work to do on navigational queries for graphs! Other lines of research with open questions: ◮ Nested regular expressions and nSPARQL [PAG09,BPR12] ◮ Queries that can output paths (e.g. ECRPQs [BLHW10]) ◮ More complexity results on counting paths [LM12] What is the right language (and the right semantics) for navigating RDF graphs?
  332. 332. Outline Basics of SPARQL Syntax and Semantics of SPARQL 1.0 What is new in SPARQL 1.1 Federation: SERVICE operator Syntax and Semantics Evaluation of SERVICE queries Navigation: Property Paths Navigating graphs with regular expressions The history of paths (in SPARQL 1.1 specification) Evaluation procedures and complexity
  333. 333. Concluding remarks ◮ Federation and Navigation are fundamental features in SPARQL 1.1 ◮ They need formalization and (serious) study
  334. 334. Concluding remarks ◮ Federation and Navigation are fundamental features in SPARQL 1.1 ◮ They need formalization and (serious) study ◮ Do not runaway from Theory! it can really help (and has helped) to understand the implications of design decisions
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×