Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

LDQL: A Query Language for the Web of Linked Data

2,100 views

Published on

I used this slideset to present our research paper at the 14th Int. Semantic Web Conference (ISWC 2015). Find a preprint of the paper here:
http://olafhartig.de/files/HartigPerez_ISWC2015_Preprint.pdf

Published in: Technology
  • DOWNLOAD FULL BOOKS, INTO AVAILABLE FORMAT ......................................................................................................................... ......................................................................................................................... 1.DOWNLOAD FULL. PDF EBOOK here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... 1.DOWNLOAD FULL. EPUB Ebook here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... 1.DOWNLOAD FULL. doc Ebook here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... 1.DOWNLOAD FULL. PDF EBOOK here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... 1.DOWNLOAD FULL. EPUB Ebook here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... 1.DOWNLOAD FULL. doc Ebook here { https://tinyurl.com/y6a5rkg5 } ......................................................................................................................... ......................................................................................................................... ......................................................................................................................... .............. Browse by Genre Available eBooks ......................................................................................................................... Art, Biography, Business, Chick Lit, Children's, Christian, Classics, Comics, Contemporary, Cookbooks, Crime, Ebooks, Fantasy, Fiction, Graphic Novels, Historical Fiction, History, Horror, Humor And Comedy, Manga, Memoir, Music, Mystery, Non Fiction, Paranormal, Philosophy, Poetry, Psychology, Religion, Romance, Science, Science Fiction, Self Help, Suspense, Spirituality, Sports, Thriller, Travel, Young Adult,
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here

LDQL: A Query Language for the Web of Linked Data

  1. 1. LDQL: A Query Language for the Web of Linked Data Olaf Hartig¹ and Jorge Pérez² ¹Hasso Plattner Institute, University of Potsdam, Germany ²Department of Computer Science, Universidad de Chile
  2. 2. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 2 Linked Data Query Processing
  3. 3. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 3 Research Question What language can we use to express queries (in a declarative manner) ?
  4. 4. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 4 Existing Proposals 1/2 ● Basic Graph Patterns – Bounded method [Bouquet, Ghidini, Serafini 2009] – Navigational method [Bouquet, Ghidini, Serafini 2009] – Direct access method [Bouquet, Ghidini, Serafini 2009] – Matching-based query semantics [Hartig 2011] – Family of authoritativeness-based query semantics [Harth and Speiser 2012] – Different inference-based query semantics [Umbrich et al. 2012]
  5. 5. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 5 Existing Proposals 2/2 ● SPARQL 1.0 – Full-Web query semantics [Hartig 2012] – Family of reachability-based semantics [Hartig 2012] ● cALL-semantics, cNONE-semantics, cMATCH-semantics, … ● SPARQL 1.1 Property Paths Patterns – Context-based query semantics [Hartig and Pirrò 2015] ● NautiLOD [Fionda, Gutierrez, and Pirrò 2012] ● LDPath (no formal semantics) [Schaffert et al. 2012]
  6. 6. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 6 Limitation of Existing Proposals ● No separation of the following two tasks: – Select query-relevant regions of the Web of Linked Data – Evaluate pattern over the data in the selected regions ● For instance, impossible to express queries such as: – Take all Linked Data reached by following two foaf:knows links starting from URI u, and evaluate a given SPARQL pattern over the union of this data.
  7. 7. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 7 General Idea of LDQL Query Specification of query-relevant region ( N , Q )
  8. 8. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 8 Our Contributions ● Formal syntax and semantics of LDQL – Including some rewrite rules (i.e., equivalences) ● Study Web safeness of LDQL queries – Show a decidable syntactic property of LDQL queries for which a complete execution is feasible in practice ● Compare LDQL with existing proposals – Show that LDQL is strictly more expressive ? ? ? ? ? ?? ?
  9. 9. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 9 Existing Proposals 2/2 ● SPARQL 1.0 – Full-Web query semantics [Hartig 2012] – Family of reachability-based semantics [Hartig 2012] ● cALL-semantics, cNONE-semantics, cMATCH-semantics, … ● SPARQL 1.1 Property Paths Patterns – Context-based query semantics [Hartig and Pirrò 2015] ● NautiLOD [Fionda, Gutierrez, and Pirrò 2012] ● LDPath (no formal semantics) [Schaffert et al. 2012] √ √ √ √ √
  10. 10. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 10 Our Contributions ● Formal syntax and semantics of LDQL – Including some rewrite rules (i.e., equivalences) ● Study Web safeness of LDQL queries – Show a decidable syntactic property of LDQL queries for which a complete execution is feasible in practice ● Compare LDQL with existing proposals – Show that LDQL is strictly more expressive ? ? ? ? ? ?? ? Let's have a brief look!
  11. 11. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 11 General Idea of LDQL Query ( N , Q ) Specification of query-relevant region
  12. 12. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 12 General Idea of LDQL SPARQL 1.1 graph pattern Link Path Expression (Nested regular expressions over an alphabet of so called link patterns) Query ( N , Q ) Specification of query-relevant region
  13. 13. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 13 Link Patterns ● Each βi is either: – a URI, or – a wildcard denoted by _, or – a placeholder for the context URI denoted by + ● β3 may also be a literal ● Example: ( β1, β2, β3 ) ( +, foaf:knows, _ )
  14. 14. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 14 Link Graph ( ua, foaf:knows, ub ) ( uc, foaf:knows, ua ) doc(ua) … doc(ub) ... doc(uc) ( +, foaf:knows, _ )
  15. 15. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 15 Link Graph ( ua, foaf:knows, ub ) ( uc, foaf:knows, ua ) doc(ua) … doc(ub) ... doc(uc) ( +, foaf:knows, _ ) ( ua, foaf:knows, ub ), ub ( uc, foaf:knows, ua ), uc( ua, foaf:knows, ub ) ( uc, foaf:knows, ua )
  16. 16. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 16 Link Pattern Matching ( ua, foaf:knows, ub ) ( uc, foaf:knows, ua ) doc(ua) … doc(ub) ... doc(uc) ( +, foaf:knows, _ ) ( ua, foaf:knows, ub ), ub ( uc, foaf:knows, ua ), uc A link graph edge labeled with RDF triple (x1, x2, x3) and URI u matches a link pattern (β1, β2, β3) in the context of URI uctx if: •  i ∈ {1,2,3} such that βi = _ and xi = u, and •  i ∈ {1,2,3} (i) βi = xi , (ii) βi = _ , or (iii) βi = + and xi = uctx
  17. 17. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 17 Link Pattern Matching ( ua, foaf:knows, ub ) ( uc, foaf:knows, ua ) doc(ua) … doc(ub) ... doc(uc) ( +, foaf:knows, _ ) ( ua, foaf:knows, ub ), ub ( uc, foaf:knows, ua ), uc A link graph edge labeled with RDF triple (x1, x2, x3) and URI u matches a link pattern (β1, β2, β3) in the context of URI uctx if: •  i ∈ {1,2,3} such that βi = _ and xi = u, and •  i ∈ {1,2,3} (i) βi = xi , (ii) βi = _ , or (iii) βi = + and xi = uctx √ X
  18. 18. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 18 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q)
  19. 19. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 19 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q) Example: Take all Linked Data reached by following two foaf:knows links starting from URI u, and evaluate SPARQL pattern P over the union of this data. ( (+,foaf:knows, _) /(+,foaf:knows, _) , P ) with seeds S = {u }
  20. 20. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 20 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q) Example: Take all Linked Data reached by following two foaf:knows links starting from URI u, and evaluate SPARQL pattern P over the union of this data. ( (+,foaf:knows, _) /(+,foaf:knows, _) , P ) with seeds S = {u }
  21. 21. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 21 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q) Example: Take all Linked Data reached by following two foaf:knows links starting from URI u, and evaluate SPARQL pattern P over the union of this data. ( (+,foaf:knows, _) /(+,foaf:knows, _) , P ) with seeds S = {u }
  22. 22. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 22 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q) For every LDQL query q', there exists a semantically equivalent LDQL query q'' such that every LPE in q'' consists only of ε and lpe* and (?v, q).
  23. 23. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 23 (Algebraic) Syntax of LDQL q := (lpe, P) | (lpe an LPE, P a SPARQL pattern) (q AND q) | (q UNION q) | πV q | (V a set of variables) (SEED U q) | (U a set of URIs) (SEED ?v q) (?v a variable) Example: ( (lpe1 , P1) AND (SEED ?x q2) ) where P1 is: ((?x,p,?y) FILTER (?y > 3)) lpe := ε | lp | lpe/lpe | lpe|lpe | lpe* | [lpe] | (?v, q)
  24. 24. LDQL: A Query Language for the Web of Linked Data – Olaf Hartig and Jorge Peréz – ISWC 2015 24 Summary ● Query language for the Web of Linked Data ● Main feature: separation of ... … navigational components to select regions of the Web of Linked Data, and … query patterns to be matched in the data of the selected regions ● In the paper we study LDQL – Web-safeness property – LDQL is strictly more expressive than the major existing proposals

×