1. China 2009http://www.larkc.eu/1 语义网与本体技术系列讲座 III 专题研究 Research Topics 黄智生 Zhisheng Huang Vrije University Amsterdam The Netherlands huang@cs.vu.nl

2. China 2009http://www.larkc.eu/2 语义网与本体技术系列讲座 第一部分：导论 2009 年 9 月 9 日星期三 14 ： 00-15 ： 30 第二部分：逻辑基础 2009 年 9 月 12 日星期六 10 ： 00-11 ： 30 第三部分：专题研究 2009 年 9 月 13 日星期日 9 ： 00-10 ： 30 ------------------------------------------------------------ LarKC 人员专题讨论 2009 年 9 月 13 日星期日 14 ： 00-15 ： 30

3. China 2009http://www.larkc.eu/3 Outline 本体推理与管理 (Reasoning and Management of Ontologies) 不一致性本体的推理（ Reasoning with Inconsistent Ontologies) 海量语义数据推理 (Scalable Reasoning) 结论和讨论 (Conclusion and Discussion)

4. China 2009http://www.larkc.eu/4 Ontology Reasoning and Inconsistency Management

5. China 2009http://www.larkc.eu/5 Inconsistency and the Semantic Web The Semantic Web is characterized by scalability, distribution, and multi-authorship All these may introduce inconsistencies.

6. China 2009http://www.larkc.eu/6 Ontologies will be inconsistent Because of: mistreatment of defaults polysemy migration from another formalism integration of multiple sources … (“Semantic Web as a wake-up call for KR”)

7. China 2009http://www.larkc.eu/7 Example: Inconsistency by mistreatment of default rules MadCow Ontology Cow  Vegetarian MadCow  Cow MadCow   Eat.BrainofSheep Sheep  Animal Vegetarian   Eat.  (Animal  PartofAnimal) Brain  PartofAnimal...... theMadCow  MadCow...

8. China 2009http://www.larkc.eu/8 Example: Inconsistency through imigration from other formalism DICE Ontology Brain  CentralNervousSystem Brain  BodyPart CentralNervousSystem  NervousSystem BodyPart   NervousSystem

9. China 2009http://www.larkc.eu/9 Inconsistency and Explosion The classical entailment is explosive: P, ¬ P |= Q Any formula is a logical consequence of a contradiction. The conclusions derived from an inconsistent ontology using the standard reasoning may be completely meaningless

10. China 2009http://www.larkc.eu/10 Why DL reasoning cannot escape the explosion The derivation checking is usually achieved by the satisfiability checking.  |=     {¬  } is not satisfiable. Tableau algorithms are approaches based on the satisfiability checking  is inconsistent =>  is not satisfiable =>   {¬  } is not satisfiable.

11. China 2009http://www.larkc.eu/11 Two main approaches to deal with inconsistency Inconsistency Diagnosis and Repair Ontology Diagnosis(Schlobach and Cornet 2003) Reasoning with Inconsistency Paraconsistent logics Limited inference (Levesque 1989) Approximate reasoning(Schaerf and Cadoli 1995) Resource-bounded inferences(Marquis et al.2003) Belief revision on relevance (Chopra et al. 2000)

12. China 2009http://www.larkc.eu/12 What an inconsistency reasoner is expected Given an inconsistent ontology, return meaningful answers to queries. General solution: Use non-standard reasoning to deal with inconsistency  |=  : the standard inference relations  |  : nonstandard inference relations

13. China 2009http://www.larkc.eu/13 Reasoning with inconsistent ontologies: Main Idea Starting from the query, 1.select consistent sub-theory by using a relevance-based selection function. 2.apply standard reasoning on the selected sub-theory to find meaningful answers. 3.If it cannot give a satisfying answer, the selection function would relax the relevance degree to extend consistent sub-theory for further reasoning.

14. China 2009http://www.larkc.eu/14 New formal notions are needed New notions: Accepted: Rejected: Overdetermined: Undetermined: Soundness: (only classically justified results) Meaningfulness: (sound & never overdetermined) soundness +

15. China 2009http://www.larkc.eu/15 Soundness:  |  =>  `  (  ` consistent and  `|=  ). Meaningfulness: sound and consistent (  |  =>  ¬  ). Local Completeness w.r.t a consistent  `  : (  `|=  =>  |  ). Maximality: locally complete w.r.t a maximal consistent set  `. Local Soundness w.r.t.a consistent set  `:  |  =>  `|=  ). Some Formal Definitions

16. China 2009http://www.larkc.eu/16 Selection Functions Given an ontology T and a query , a selection function s(T, ,k) returns a subset of the ontology at each step k>0.

17. China 2009http://www.larkc.eu/17 General framework Use selection function s(T, ,k), with s(T, ,k)  s(T, ,k+1) 1.Start with k=0: s(T, ,0) |=  or s(T, ,0) |=  ? 2.Increase k, until s(T, ,k) |=  or s(T, ,k) |=  3.Abort when undetermined at maximal k overdetermined at some k

18. China 2009http://www.larkc.eu/18 Inconsistency Reasoning Processing: Linear Extension

19. China 2009http://www.larkc.eu/19 Proposition: Linear Extension Never over-determined May undetermined Always sound Always meaningful Always locally complete May not maximal Always locally sound

20. China 2009http://www.larkc.eu/20 Direct Relevance and K Relevance Direct relevance (0-relevance). there is a common name in two formulas: C(  )  C(  )   R(  )  R(  )   I(  )  I(  ) . K-relevance: there exist formulas  0,  1,…,  k such that  and  0,  0 and  1, …,  k and  are directly relevant.

21. China 2009http://www.larkc.eu/21 Relevance-based Selection Functions s(T, ,0)=  s(T, ,1)= {  T:  is directly relevant to  }. s(T, ,k)= {  T:  is directly relevant to s(T, ,k-1)}.

22. China 2009http://www.larkc.eu/22 PION Prototype PION: Processing Inconsistent ONtologies http://wasp.cs.vu.nl/sekt/pion

23. China 2009http://www.larkc.eu/23 An Extended DIG Description Logic Interface for Prolog (XDIG) A logic programming infrastructure for the Semantic Web Similar to SOAP Application independent, platform independent Support for DIG clients and DIG servers.

24. China 2009http://www.larkc.eu/24 XDIG As a DIG client, the Prolog programs can call any external DL reasoner which supports the DIG DL interface. As a DIG server, the Prolog programs can serve as a DL reasoner, which can be used to support additional reasoning processing, like inconsistency reasoning multi-version reasoning, and inconsistency diagnosis and repair.

25. China 2009http://www.larkc.eu/25 XDIG package The XDIG package and the source code are now available for public download at the website: http://wasp.cs.vu.nl/sekt/dig/ In the package, we offer five examples how XDIG can be used to develop extended DL reasoners.

26. China 2009http://www.larkc.eu/26 PION Testbed

27. China 2009http://www.larkc.eu/27 Answer Evaluation Intended Answer (IA): PION answer = Intuitive Answer Cautious Answer (CA): PION answer is ‘undetermined’, but intuitive answer is ‘accepted’ or ‘rejected’. Reckless Answer (RA): PION answer is ‘accepted’ or ‘rejected’, but intuitive answer is ‘undetermined’. Counter Intuitive Answer (CIA): PION answer is ‘accepted’ but intuitive answer is ‘rejected’, or vice verse.

28. China 2009http://www.larkc.eu/28 Preliminary Tests with Syntactic-relevance Selection Function OntologyQueriesIACARACIAIA (%) ICR (%) Bird50 000100 Brain (DICE) 423642085.7100 Married Woman 504802096100 MadCow254236160292.999

29. China 2009http://www.larkc.eu/29 Intensive Tests on PION Evaluation and test on PION with several realistic ontologies: Communication Ontology Transportation Ontology MadCow Ontology Each ontology has been tested by thousands of queries with different selection functions.

30. China 2009http://www.larkc.eu/30 Summary we proposed a general framework for reasoning with inconsistent ontologies based on selecting ever increasing consistent subsets choice of selection function is crucial query-based selection functions are flexible to find intended answers simple syntactic selection works surprisingly well

31. China 2009http://www.larkc.eu/31 Extension Semantic Relevance Based Selection Functions K-extension Variants of over-determined processing strategies Integrating with the diagnosis approach

32. http://www.larkc.eu/32 Using Semantic Distances for Reasoning with Inconsistent Ontologies Google distances are used to develop semantic relevance functions to reason with inconsistent ontologies. Assumption: two concepts appear more frequently in the same web page, they are semantically more similar (relevant).

33. http://www.larkc.eu/33 Google Distances (Cilibrasi and Vitanyi 2004) Google distance is measured in terms of the co- occurrence of two search items in the Web by Google search engine. Normalized Google Distance (NGD) is introduced to measure the similarity/light-weight semantic relevance NGD(x,y)= (max{log f(x), log f(y)}-log f(x,y))/(log M- min{log f(x),log f(y)} where f(x) is the number of Google hits for x f(x,y) is the number of Google hits for the tuple of search items x and y M is the number of web pages indexed by Google.

34. China 2009http://www.larkc.eu/34 Semantic Distances Define semantic distances (SD) between two formulas in terms of semantic distances between two concepts/roles/individuals (NGD)

35. China 2009http://www.larkc.eu/35 Postulates for Semantic Distances

36. http://www.larkc.eu/36 Semantic Distances Semantic distance are measured by the ratio of the summed distance of the difference between two formulae to the maximal distance between two formulae.

37. China 2009http://www.larkc.eu/37 Proposition The semantic distance SD satisfies the properties Range, Reflexivity, Symmetry, Maximum Distance, and Intermediate Values.

38. China 2009http://www.larkc.eu/38 Example: MadCow NGD(MadCow, Grass)=0.7229 NGD(MadCow, Sheep)=0.6120

39. China 2009http://www.larkc.eu/39 Implementation: PION PION: Processing Inconsistent ONtologies http://wasp.cs.vu.nl/sekt/pion

40. http://www.larkc.eu/40 Answer Evaluation Intended Answer (IA): Query answer = Intuitive Answer Cautious Answer (CA): Query answer is ‘undetermined’, but Intutitve answer is ‘accepted’ or ‘rejected’. Reckless Answer (RA): Query answer is ‘accepted’ or ‘rejected’, but Intutive answer is ‘undetermined’. Counter Intuitive Answer (CIA): Query answer is ‘accepted’ but Intuitive answer is ‘rejected’, or vice versa.

41. http://www.larkc.eu/41 Syntactic approach vs. Semantic approach: quality of query answers

42. http://www.larkc.eu/42 Syntactic approach vs. Semantic approach: Time Performance

43. http://www.larkc.eu/43 Summary The run-time of the semantic approach is much better than the syntactic approach, while the quality remains comparable. The semantic approach can be parameterised so as to stepwise further improve the run-time with only a very small drop in quality.

44. http://www.larkc.eu/44 Summary (cont.) The semantic approach for reasoning with inconsistent ontologies trade-off computational cost for inferential completeness, and provide attractive scalability.

45. China 2009http://www.larkc.eu/45 LarKC: 一个海量语义数据处理 平台 The Large Knowledge Collider ( 大型知识 对撞机） A configurable platform for experimentation by others http://www.larkc.eu

46. China 2009http://www.larkc.eu/46 可布局平台 “Configurable platform” “a configurable platform for infinitely scalable semantic web reasoning”. Enrich current logic-based Semantic Web reasoning with methods from information retrieval, machine learning, information theory, databases, and probabilistic reasoning

47. China 2009http://www.larkc.eu/47 网络科学与人类智能科学的结合 Web Science with Human Intelligence Employing cognitively inspired approaches and techniques such as spreading activation, focus of attention, reinforcement, habituation, relevance reasoning, and bounded rationality

48. China 2009http://www.larkc.eu/48 Achieve scalability through giving up completeness by giving up 100% correctness: trading quality for size often completeness is not needed sometimes even correctness is not needed precision (soundness) recall (completeness) logic IR Semantic Web

49. China 2009http://www.larkc.eu/49 通过并行计算达到海量数据处理能力 Achieve Scalability through Parallelization by parallelisation: cluster computing wide area distribution “Thinking@home”, “self-computing semantic Web” cloud computing 云计算 (Amazon ， Google)

50. China 2009http://www.larkc.eu/50 欧盟第七框架研究课题 : LarKC EU 7th framework Project 总预算 1 千万欧元： 10M€ budget 历时 3 年半： 3.5 years 八十个人年： 80 person years 3 个实例研究： 3 case studies 14 个合作单位： 14 partners, 来自 12 个国家： 12 countries, 来自 3 大洲： 3 continents project nr. FP7 – 215535

51. China 2009http://www.larkc.eu/51 The consortium 50 people present

52. China 2009http://www.larkc.eu/52 The Consortium Combining consortium competence IR, Cognition ML, Ontologies Statistics, ML, Cognition,DB Logic,DB, Probabilistic Inference Economics, Decision Theory

53. China 2009http://www.larkc.eu/53 课题组成 Project Workpackages WP1 – Conceptual Framework & Evaluation WP 2: Retrieval and Selection WP5: Collider Platform WP 9: Exploitation and standards WP 10: Project Management WP 8: Training, dissemination, community building WP3: Abstraction and Learning WP4: Reasoning and Deciding WP 6: Use case: Real Time City WP 7a: Use case: Early Clinical Development WP 7b: Use case: Carcinogenesis Reference Production

54. China 2009http://www.larkc.eu/54 Use case: Drug Discovery Problem: pharmaceutical R&D in early clinical development is stagnating (Q1Q2Q3)(Q1Q2Q3) FDA white paper Innovation or Stagnation (March 2004): “developers have no choice but to use the tools of the last century to assess this century's candidate solutions.” “industry scientists often lack cross-cutting information about an entire product area, or information about techniques that may be used in areas other than theirs” FDA white paper Innovation or Stagnation (March 2004): “developers have no choice but to use the tools of the last century to assess this century's candidate solutions.” “industry scientists often lack cross-cutting information about an entire product area, or information about techniques that may be used in areas other than theirs” “ Show me any potential liver toxicity associated with the compound’s drug class, target, structure and disease.” Show me all liver toxicity associated with the target or the pathway. Genetics “Show me all liver toxicity associated with compounds with similar structure” Chemistry “Show me all liver toxicity from the public literature and internal reports that are related to the drug class, disease and patient population” LITERATURE Current NCBI: linking but no inference

55. China 2009http://www.larkc.eu/55 Use Case: Real Time City Our cities face many challenges Urban Computing is the ICT way to address them How can we redevelop existing neighborhoods and business districts to improve the quality of life? How can we create more choices in housing, accommodating diverse lifestyles and all income levels? How can we reduce traffic congestion yet stay connected? How can we include citizens in planning their communities rather than limiting input to only those affected by the next project? How can we fund schools, bridges, roads, and clean water while meeting short-term costs of increased security? How can we redevelop existing neighborhoods and business districts to improve the quality of life? How can we create more choices in housing, accommodating diverse lifestyles and all income levels? How can we reduce traffic congestion yet stay connected? How can we include citizens in planning their communities rather than limiting input to only those affected by the next project? How can we fund schools, bridges, roads, and clean water while meeting short-term costs of increased security? Is public transportation where the people are? Which landmarks attract more people? Where are people concentrating? Where is traffic moving?

56. China 2009http://www.larkc.eu/56 课题时间表 Project Timeline Surveys (plugins, platform) Requirements (use cases) Prototype Internal Release Public Release Final Release Use Cases V1 Use Cases V2 Use Cases V3 4206183310

57. China 2009http://www.larkc.eu/57 如果你对参与开发感兴趣的话 How can any other interested party contribute? The Large Knowledge Collider is an open, and configurable platform. The first public version of the Large Knowledge Collider is available. LarKC has formed an "early adapters group". LarKC will actively support this group in use the Large Knowledge Collider platform.

58. China 2009http://www.larkc.eu/58 LarKC 中文论坛 http://groups.google.com/group/larkc- chinese-forum

59. China 2009http://www.larkc.eu/59 Realising the Architecture Pipeline Support System Pipeline Support System Plug-in Registry Plug-in Manager Data Layer Plug-in API Data Layer API RDF Store RDF Store 59

60. China 2009http://www.larkc.eu/60 Data Layer API Pipeline Support System Pipeline Support System Plug-in Registry RDF Store RDF Store RDF Store RDF Store RDF Store RDF Store RDF Doc RDF Doc RDF Doc RDF Doc Data Layer Decider Plug-in API Plug-in Manager Query Transformer Query Transformer Plug-in API Plug-in Manager Identifier Plug-in API Plug-in Manager Info. Set Transformer Info. Set Transformer Plug-in API Plug-in Manager Selecter Plug-in API Plug-in Manager Reasoner Plug-in API Application RDF Doc RDF Doc Platform Utility Functionality APIs Plug-ins External systems External data sources LarKC Architecture 60

61. China 2009http://www.larkc.eu/61 LarKC Plug-in API: General Plug-in Model Plug-ins are identified by a URI (Uniform Resource Identifier) Plug-ins provide MetaData about what they do (Functional properties): e.g. type = Selecter Plug-ins provide information about their behaviour and needs, including Quality of Service information (Non-functional properties): e.g. Throughput, MinMemory, Cost,… + URI getIdentifier() + QoSInformation getQoSInformation() + URI getIdentifier() + QoSInformation getQoSInformation() Plug-in  Functional properties  Non-functional properties  WSDL description  Functional properties  Non-functional properties  WSDL description Plug-in description 61

62. China 2009http://www.larkc.eu/62 LarKC Plug-in API: IDENTIFY IDENTIFY: Given a query, identify resources that could be used to answer it Sindice – Triple Pattern Query  RDF Graphs Google – Keyword Query  Natural Language Document Triple Store – SPARQL Query  RDF Graphs + Collection identify (Query theQuery, Contract contract, Context context) + Collection identify (Query theQuery, Contract contract, Context context) Identifier 62

63. China 2009http://www.larkc.eu/63 LarKC Plug-in API: TRANSFORM (1/2) Query TRANSFORM: Transforms a query from one representation to another SPARQL Query  Triple Pattern Query SPARQL Query  Keyword Query SPARQL Query  SPARQL Query (different abstraction) SQARQL Query  CycL Query + Set transform(Query theQuery, Contract theContract, Context theContext) QueryTransformer 63

64. China 2009http://www.larkc.eu/64 LarKC Plug-in API: TRANSFORM (2/2) Information Set TRANSFORM: Transforms data from one representation to another Natural Language Document  RDF Graph Structured Data Sources  RDF Graph RDF Graph  RDF Graph (e.g. foaf vocabulary to facebook vocabulary ) + InformationSet transform(InformationSet theInformationSet, Contract theContract, Context theContext) InformationSetTransformer 64

65. China 2009http://www.larkc.eu/65 LarKC Plug-in API: SELECT SELECT: Given a set of statements (e.g. a number of RDF Graphs) will choose a selection/sample from this set Collection of RDF Graphs  Triple Set (Merged) Collection of RDF Graphs  Triple Set (10% of each) Collection of RDF Graphs  Triple Set (N Triples) + SetOfStatements select(SetOfStatements theSetOfStatements, Contract contract, Context context) + SetOfStatements select(SetOfStatements theSetOfStatements, Contract contract, Context context) Selecter 65

66. China 2009http://www.larkc.eu/66 LarKC Plug-in API: REASON REASON: Executes a query against the supplied set of statements SPARQL Query  Variable Binding (Select) SPARQL Query  Set of statements (Construct) SPARQL Query  Set of statements (Describe) SPARQL Query  Boolean (Ask) + VariableBinding sparqlSelect(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + SetOfStatements sparqlConstruct(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + SetOfStatements sparqlDescribe(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + BooleanInformationSet sparqlAsk(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + VariableBinding sparqlSelect(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + SetOfStatements sparqlConstruct(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + SetOfStatements sparqlDescribe(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) + BooleanInformationSet sparqlAsk(SPARQLQuery theQuery, SetOfStatements theSetOfStatements, Contract contract, Context context) Reasoner 66

67. China 2009http://www.larkc.eu/67 LarKC Plug-in API: DECIDE DECIDE: Builds the pipeline and manages the control flow Scripted Decider: Predefined pipeline is built and executed Self-configuring Decider: Uses plug-in descriptions (functional and non- functional properties) to build the pipeline + VariableBinding sparqlSelect(SPARQLQuery theQuery, QoSParameters theQoSParameters) + SetOfStatements sparqlConstruct(SPARQLQuery theQuery, QoSParameters theQoSParameters) + SetOfStatements sparqlDescribe(SPARQLQuery theQuery, QoSParameters theQoSParameters) + BooleanInformationSet sparqlAsk(SPARQLQuery theQuery, QoSParameters theQoSParameters) + VariableBinding sparqlSelect(SPARQLQuery theQuery, QoSParameters theQoSParameters) + SetOfStatements sparqlConstruct(SPARQLQuery theQuery, QoSParameters theQoSParameters) + SetOfStatements sparqlDescribe(SPARQLQuery theQuery, QoSParameters theQoSParameters) + BooleanInformationSet sparqlAsk(SPARQLQuery theQuery, QoSParameters theQoSParameters) Decider 67

68. China 2009http://www.larkc.eu/68 Decider Plug-in API Plug-in Manager Query Transformer Query Transformer Plug-in API Plug-in Manager Identifier Plug-in API Plug-in Manager Info. Set Transformer Info. Set Transformer Plug-in API Plug-in Manager Selecter Plug-in API Plug-in Manager Reasoner Plug-in API Plug-in Registry Pipeline Support System Pipeline Support System RDF Store RDF Store Identifier Info Set Transformer Reasoner Decider Selecter Query Transformer Query Transformer What does a pipeline look like? 68

69. China 2009http://www.larkc.eu/69 LarKC Data Model :Transport By Reference RDF Graph Default Graph RDF Graph Dataset: Collection of named graphs Labeled Set: Pointers to data Current Scale: O(10 10 ) triples 69

70. China 2009http://www.larkc.eu/70 LarKC Platform and the DIG plug-in LarKC Platform DIG Interface Plug-in Racer FACT++ KAON2

71. China 2009http://www.larkc.eu/71 Tasks of the DIG Plug-in 1.Translate a set of statements (ontology data) into a DIG data. If it is OWL-DL data, the use the OWL2DIG library to translate it into a DIG data 2.Translate SPARQL(DL) query into DIG - deal with triple-encoded DL expressions 3. Query processing and answer checking 4. Translate DIG answers into SPARQL answers 71 footer 23/06/2015

72. China 2009http://www.larkc.eu/72 LarKC Platform and the DIG plug-in LarKC Platform DIG Interface Plug-in External DIG Reasoner Ontology (URI)/ Set of Statements Tell SPARQL query Ask Response SPARQL Answer

73. China 2009http://www.larkc.eu/73 The DIG plug-in (v0.3) Have been supported Support the DIG interface 1.1. Support Sparqlask and Sparqlselect. DL Expressions (conjunction, disjunction, disjoint, negation) DIG queries (subsumption, instance, instances) Have been tested with Racer1.7.14 PION 2.1.0 To be supported soon: Complex DL concept expressions (such as nominal, min, max, etc.) Complex Sparql expressions (such as Filtering, Optional, Regular expressions, sparqlconstruct, sparqldescribe, etc.) Complex DIG queries (role query, functional query, value pair query)

74. China 200974 Why SPARQL-DL? SPARQL is too expressive for a DL reasoner can support. In SPARQL, there is no semantic interpretation for DL expressions such as owl:sameas, owl:disjointwith, etc. SPARQL-DL is a DL-specific SPARQL with some DL primitives, such as type(a, C), SubClassof(C1, C2), DisjointWith(C1,C2), ComplementOf(C1,C2),EquivalentClass( C1,C2),…(Sirin and Parsia 2007)

75. China 2009http://www.larkc.eu/75 Translation of DL expressions into RDF triples Using the OWL-DL method (Patel- Schneider,Hayes, Horrocks 2004). http://www.w3.org/TR/owl- semantics/mapping.html

76. China 2009http://www.larkc.eu/76 SPARQL-DL Query Example 1 ?- subClassOf(Wine, PotableLiquid) // to ask whether or not wine is a subclass of potable liquid PREFIX rdfs: PREFIX wine: PREFIX food: ASK WHERE { wine:Wine rdfs:subClassOf food:PotableLiquid.}

77. China 2009http://www.larkc.eu/77 SPARQL-DL Query Example 2 ?- subClassOf(Bordeaux, and(SweetWine, TableWine)) // to ask whether or not Bordeaux is a SweetWine and TableWine PREFIX rdfs: PREFIX rdf: …… PREFIX owl: ASK wine:Bordeaux rdfs:subClassOf _:x. _:x owl:interSectionOf _:y1. _:y1 rdf:first wine:SweetWine. _:y1 rdf:rest wine:TableWine. wine:Bordeaux rdf:type owl:Class.}

78. China 2009http://www.larkc.eu/78 Simple SPARQLSelect Query: Example 3 ?- subClassOf(?X, Wine) // to list all subconcepts of Wine PREFIX rdfs: PREFIX wine: SELECT ?X WHERE { ?X rdfs:subClassOf wine:Wine.}

79. China 2009http://www.larkc.eu/79 SPARQL-DL Query Example 4 ?- subClassOf(Bordeaux, ?X), subClassOf(?X,Wine), subClassOf(?X,?Y). PREFIX rdfs:http://www.w3.org/2000/01/rdf-schema#.. ….. PREFIX wine: SELECT ?X ?Y WHERE { wine:Bordeaux rdfs:subClassOf ?X. ?X rdfs:subClassOf wine:Wine. ?X rdfs:subClassOf ?Y. ?Y rdf:type owl:Class.}

80. China 2009http://www.larkc.eu/80 PION and External DIG Reasoner PION needs an external DIG Reasoner for standard reasoning(i.e., non-inconsistency reasoning)

81. China 2009http://www.larkc.eu/81 Compare it with that from a standard DL reasoner You can see that when querying an inconsistent ontology, the standard DL reasoner always returns an error message, like this: <responses xmlns="http://dl.kr.org/dig/2003/02/lang" xmlns:xsi="http://www.w3.org/2001/XMLSchema- instance" xsi:schemaLocation="http://dl.kr.org/dig/2003/02/lang http://dl-web.man.ac.uk/dig/2003/02/dig.xsd"> <error id="http://wasp.cs.vu.nl/larkc/ontology/ex#themadcow http://www.w3.org/1999/02/22-rdf-syntax- ns#type http://wasp.cs.vu.nl/larkc/ontolog/ex#vegetarian" message="ABox http://dl.kr.org/dig/kb-1048 is incoherent."/>

82. China 2009http://www.larkc.eu/82 Various Strategies You can use the PION testbed page piontest2.htm to select different strategies for reasoning with inconsistent ontologies by PION: selection functions (syntactic relevance, concept syntactic relevance, or semantic relevance by Google distances), over-determed processing methods (first maximal consistent set, or path pruning with Google distances), extension strategies (linear extension or k- extension).

83. China 2009http://www.larkc.eu/83 Questions and Discussions