Presentation is loading. Please wait.

Presentation is loading. Please wait.

ISO TC37/SC4 N429 ISO/TC37/SC4/TDG6 Language Resource Ontologies 2007-11-11/12, Busan 2007-11-11/12, Busan HASIDA Koiti HASIDA Koiti

Similar presentations


Presentation on theme: "ISO TC37/SC4 N429 ISO/TC37/SC4/TDG6 Language Resource Ontologies 2007-11-11/12, Busan 2007-11-11/12, Busan HASIDA Koiti HASIDA Koiti"— Presentation transcript:

1 ISO TC37/SC4 N429 ISO/TC37/SC4/TDG6 Language Resource Ontologies 2007-11-11/12, Busan 2007-11-11/12, Busan HASIDA Koiti HASIDA Koiti hasida.k@aist.go.jp hasida.k@aist.go.jp ITRI, AIST, Japan ITRI, AIST, Japan

2 Agenda 1st Day (Sunday, November 11, 2007) 1st Day (Sunday, November 11, 2007) 16:30/16:45 Introduction to ISO/TC37/SC4/TDG6 (Koiti Hasida) 16:30/16:45 Introduction to ISO/TC37/SC4/TDG6 (Koiti Hasida) 16:45/18:00 RDF Issues (Koiti Hasida, etc.) 16:45/18:00 RDF Issues (Koiti Hasida, etc.) 18:00/18:15 break 18:00/18:15 break 18:15/20:00 Ontology Issues (Paul Buitelaar, Aldo Gangemi, etc.) 18:15/20:00 Ontology Issues (Paul Buitelaar, Aldo Gangemi, etc.) (Dinner will be provided for the participants) (Dinner will be provided for the participants) 2nd Day (Monday, November 12, 2007) 2nd Day (Monday, November 12, 2007) 15:00/16:30 Ontology of Taxonomies (Bodil Nistrup Madsen) 15:00/16:30 Ontology of Taxonomies (Bodil Nistrup Madsen) 16:30/16:40 break 16:30/16:40 break 16:40/17:20 Ontology Work Guide for Web Content (Miran Choi) 16:40/17:20 Ontology Work Guide for Web Content (Miran Choi) 17:20/18:00 Wrap-Up 17:20/18:00 Wrap-Up 2

3 ISO/TC37 Terminology and Other Language and Content Resources SC1: Principles and Methods SC1: Principles and Methods SC2: Terminography and Lexicography SC2: Terminography and Lexicography SC3: Computer Applications for Terminology SC3: Computer Applications for Terminology ISO12620: Data Categories ISO12620: Data Categories SC4: Language Resources Management SC4: Language Resources Management Technical Committee Intl. Standardization Org. Sub-Committee 3

4 ISO/TC37/SC4 WG1 Basic Description and Mechanisms for Language Resources (Laurent Romary) WG1 Basic Description and Mechanisms for Language Resources (Laurent Romary) ISO 24610-1:2006 Feature Structure - Part 1:Feature System Representation (FSR) Kiyong Lee ISO 24610-1:2006 Feature Structure - Part 1:Feature System Representation (FSR) Kiyong Lee ISO 24610-2 Feature Structure-Part 2: Feature System Declaration (FSD) Kiyong Lee ISO 24610-2 Feature Structure-Part 2: Feature System Declaration (FSD) Kiyong Lee ISO 24612 Linguistic Annotation Framework (LAF) Nancy Ide ISO 24612 Linguistic Annotation Framework (LAF) Nancy Ide ISO 24618 Citation of Electronic Resource (CitER) Daan Broeder ISO 24618 Citation of Electronic Resource (CitER) Daan Broeder WG2 Representation Schemes (Kiyong Lee) WG2 Representation Schemes (Kiyong Lee) ISO24614-1 Word Segmentation - Part1: Basic concepts and General Principle (WordSeg-1) Maosong Sun ISO24614-1 Word Segmentation - Part1: Basic concepts and General Principle (WordSeg-1) Maosong Sun ISO 24614-2 Word Segmentation - Part 2: Chinese, Japanese and Korean (WordSeg-2) ISO 24614-2 Word Segmentation - Part 2: Chinese, Japanese and Korean (WordSeg-2) ISO 24617-1 Semantic Annotation Framework - Part1: Time and Events (SemAF/Time) ISO 24617-1 Semantic Annotation Framework - Part1: Time and Events (SemAF/Time) ISO 24615 Syntactic Annotation Framework (SynAF) Thierry Declerck ISO 24615 Syntactic Annotation Framework (SynAF) Thierry Declerck ISO 24611 Morpho-Syntactic Annotation Framework (MAF) ISO 24611 Morpho-Syntactic Annotation Framework (MAF) WG3 Multilingual Information Representation WG3 Multilingual Information Representation ISO 24616-1 Multi-lingual information framework (MLIF) ISO 24616-1 Multi-lingual information framework (MLIF) WG4 Lexical Resources/database (Nicoretta Calzolari) WG4 Lexical Resources/database (Nicoretta Calzolari) ISO 24613 Lexical Markup Framework ISO 24613 Lexical Markup Framework TDG1 Data Categories for Metadata (Peter Wittenburg) TDG1 Data Categories for Metadata (Peter Wittenburg) TDG2 Data Categories for Morphsyntax (Gil Francopoulo) TDG2 Data Categories for Morphsyntax (Gil Francopoulo) TDG3 Data Categories for Semantic Content Representation (Harry Bunt) TDG3 Data Categories for Semantic Content Representation (Harry Bunt) TDG4 Syntactic Data Categories (Thierry Declerck) TDG4 Syntactic Data Categories (Thierry Declerck) TDG5 Machine Readable Dictionaries (Monte George) TDG5 Machine Readable Dictionaries (Monte George) TDG6 Language Resource Ontologies (Koiti Hasida) TDG6 Language Resource Ontologies (Koiti Hasida) 4 Thematic Domain Grp.

5 TDG6 Issues ontologization ontologization DC, LAF, LMF, FS, MAF, SemAF, SynAF, TDG3, etc. DC, LAF, LMF, FS, MAF, SemAF, SynAF, TDG3, etc. Cf. the Pisa group’s work on LMF Cf. the Pisa group’s work on LMF extension of RDF (and ontology framework) to more straightforwardly address linguistic information extension of RDF (and ontology framework) to more straightforwardly address linguistic information extended RDF instead of XML extended RDF instead of XML nodes embedding nodes … rdf:Container? nodes embedding nodes … rdf:Container? publish TRs publish TRs launch ISs launch ISs 5

6 Ontologization ontology-based reformulation ontology-based reformulation Most current standards are based on XML and lack formal metamodeling tools. Most current standards are based on XML and lack formal metamodeling tools. not XML but RDF as base description and modeling tool not XML but RDF as base description and modeling tool Semantic interpretation is standardized not for XML but for RDF. Semantic interpretation is standardized not for XML but for RDF. something like N3 for textual representation something like N3 for textual representation ontology as schema ontology as schema not DTD, XML Schema, RELAXNG, etc. not DTD, XML Schema, RELAXNG, etc. 6

7 Motivations of Ontologization Lack of formal tool by which to write schemas fully addressing the specifications. Lack of formal tool by which to write schemas fully addressing the specifications. DCR model lacks descriptive power. DCR model lacks descriptive power. 7

8 Semantic Mess of XML Semantic interpretation of XML is not standardized but rather arbitrary. Semantic interpretation of XML is not standardized but rather arbitrary. Many inconsistent `standards’ on overlapping issues. Many inconsistent `standards’ on overlapping issues. Huge standards containing many different semantic interpretation manners. Huge standards containing many different semantic interpretation manners. e.g., MPEG-7 > 2000 pages e.g., MPEG-7 > 2000 pages 8

9 RDF Resource Description Framework Resource Description Framework W3C recommendation http://www.w3.org/RDF/ W3C recommendation http://www.w3.org/RDF/ http://www.w3.org/RDF/ basis of ontology standards such as RDFS, OWL, and SKOS. basis of ontology standards such as RDFS, OWL, and SKOS. graph model graph model textual representation textual representation XML XML N3 N3 9

10 RDF Graph http://www.example.org/people#fredhttp://www.example.org/people#fred http://meetings.example.com/m1/hphttp://meetings.example.com/m1/hp m:homePagem:homePage m:attendingm:attending http://meetings.example.com/cal#m1http://meetings.example.com/cal#m1 m:givenNamem:givenName FredFred m:hasEmailm:hasEmail mailto:fred@example.commailto:fred@example.com 10

11 RDF in Text XML N3 <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:m="http://www.example.org/meeting_organization#" xmlns="http://www.example.org/people#" xmlns:p="http://www.example.org/personal_details#"> Fred <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:m="http://www.example.org/meeting_organization#" xmlns="http://www.example.org/people#" xmlns:p="http://www.example.org/personal_details#"> Fred @prefix p:. @prefix m:. m:homePage. p:GivenName "Fred"; p:hasEmail ; m:attending. @prefix p:. @prefix m:. m:homePage. p:GivenName "Fred"; p:hasEmail ; m:attending. 11

12 <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:m="http://www.example.org/meeting_organization#" xmlns="http://www.example.org/people#" xmlns:p="http://www.example.org/personal_details#"> Fred <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:m="http://www.example.org/meeting_organization#" xmlns="http://www.example.org/people#" xmlns:p="http://www.example.org/personal_details#"> Fred http://www.example.org/people#fredhttp://www.example.org/people#fred http://meetings.example.com/m1/hphttp://meetings.example.com/m1/hp m:homePagem:homePage m:attendingm:attending http://meetings.example.com/cal#m1http://meetings.example.com/cal#m1 m:givenNamem:givenName FredFred m:hasEmailm:hasEmail mailto:fred@example.commailto:fred@example.com 12

13 http://www.example.org/people#fredhttp://www.example.org/people#fred http://meetings.example.com/m1/hphttp://meetings.example.com/m1/hp m:homePagem:homePage m:attendingm:attending http://meetings.example.com/cal#m1http://meetings.example.com/cal#m1 m:givenNamem:givenName FredFred m:hasEmailm:hasEmail mailto:fred@example.commailto:fred@example.com @prefix p:. @prefix m:. m:homePage. p:GivenName "Fred"; p:hasEmail ; m:attending. @prefix p:. @prefix m:. m:homePage. p:GivenName "Fred"; p:hasEmail ; m:attending. 13

14 ISO 24610: Feature Structure typed feature structure as in HPSG, etc. typed feature structure as in HPSG, etc. ISO 24610-1: Feature Structure Representation ISO 24610-1: Feature Structure Representation cf. ISO 24610-2: Feature System Declaration cf. ISO 24610-2: Feature System Declaration graph model graph model AVM (attribute-value matrix) AVM (attribute-value matrix) textual encoding textual encoding XML XML 14

15 FS Graph determinerdeterminer POSPOS SPECIFIERSPECIFIER ORTHORTHlala HEADHEAD AGRAGR AGRAGR nounnoun POSPOS ORTHORTH pommepomme singularsingular NUMBERNUMBER 15

16 FS in AVM SPECIFIER HEAD POSdeterminer ORTH`la’ AGR [1] [NUMBER singular] POSnoun ORTH`pomme’ AGR [1] 16

17 FS in XML la pomme la pomme 17

18 FS in RDF (i.e., Graph) determinerdeterminer POSPOS SPECIFIERSPECIFIER ORTHORTHlala HEADHEAD AGRAGR AGRAGR nounnoun POSPOS ORTHORTHpommepomme singularsingular NUMBERNUMBER 18

19 FS in RDF/N3 [ fs:specifier [ fs:pos fs:determiner; fs:orth "la"; fs:agr ] fs:head [ fs:pos fs:noun; fs:orth "pomme"; fs:agr ] ] fs:number fs:singular. [ fs:specifier [ fs:pos fs:determiner; fs:orth "la"; fs:agr ] fs:head [ fs:pos fs:noun; fs:orth "pomme"; fs:agr ] ] fs:number fs:singular. 19

20 Ontologies Subsume Feature Systems Features are partial functions, whereas RDF properties are mappings in general (possibly partial functions). Features are partial functions, whereas RDF properties are mappings in general (possibly partial functions). Usual feature systems have no taxonomy of features, whereas usual ontologies have taxonomies of properties (e.g., due to rdfs:subPropertyOf). Usual feature systems have no taxonomy of features, whereas usual ontologies have taxonomies of properties (e.g., due to rdfs:subPropertyOf). 20

21 Embedded Situation in N3 Formula Tom believes that Mary kisses Bill. Tom believes that Mary kisses Bill. [a :believe; :experiencer ; :content { a :kiss; :agent ; :patient } ] [a :believe; :experiencer ; :content { a :kiss; :agent ; :patient } ] 21

22 FS Ontologization (Summary) RDF ⊃ FS RDF ⊃ FS Use N3 (instead of XML) for textual representation of FS as a kind of RDF. Use N3 (instead of XML) for textual representation of FS as a kind of RDF. Use ontologies for feature-system declarations. Use ontologies for feature-system declarations. 22

23 XML vs. RDF <syn:S> Tom Tom doesn’t look doesn’t look good. good.</syn:S> ● doesn’t look ●. TomTom goodgood dict1:gooddict1:good semsem a node embedding nodes S NP AJP 23

24 RDF in Graph vs. N3 24 [a syn:S; syn:child [a syn:NP; syn:child ”Tom”]; syn:child ” doesn’t look ”; syn:child [a syn:AJP; :sem dict1:good; syn:child ”good”]; syn:child ”.” ] [a syn:S; syn:child [a syn:NP; syn:child ”Tom”]; syn:child ” doesn’t look ”; syn:child [a syn:AJP; :sem dict1:good; syn:child ”good”]; syn:child ”.” ] optional and ordered ● doesn’t look ●. TomTom goodgood dict1:gooddict1:good :sem:sem S NP AJP

25 XML vs. RDF/N3 25 [a syn:S; [a syn:NP; ”Tom”]; ” doesn’t look ”; [a syn:AJP; :sem dict1:good; ”good”]; ”.” ] [a syn:S; [a syn:NP; ”Tom”]; ” doesn’t look ”; [a syn:AJP; :sem dict1:good; ”good”]; ”.” ] <syn:S> Tom Tom doesn’t look doesn’t look good. good.</S><syn:S> Tom Tom doesn’t look doesn’t look good. good.</S>

26 Cf. XML vs. RDF/XML 26 ” ” Tom doesn’t look ” good. ” ” Tom doesn’t look ” good. <syn:S> Tom Tom doesn’t look doesn’t look good. good.</S><syn:S> Tom Tom doesn’t look doesn’t look good. good.</S> Of course we won’t do this.

27 Merits of N3 over XML N3 is restricted to RDF, whereas XML has unnecessary freedom. N3 is restricted to RDF, whereas XML has unnecessary freedom. It’s confusing to use XML to represent RDF encoding XML. It’s confusing to use XML to represent RDF encoding XML. N3 can readily represent embedded situations. (cf. named graph) N3 can readily represent embedded situations. (cf. named graph) 27

28 Weaknesses of DCR Metamodel DCR metamodel cannot address DCR metamodel cannot address sorts of DCs: such as unary predicate, binary relation, symmetric binary relation, etc. sorts of DCs: such as unary predicate, binary relation, symmetric binary relation, etc. types of the domain (1 st arg.) and the range (2 nd arg.) of binary relations (properties) types of the domain (1 st arg.) and the range (2 nd arg.) of binary relations (properties) 28

29 Discussion What’s essential is the graph model. What’s essential is the graph model. Textual representation is not essential. Textual representation is not essential. But people tend to specify XML representation, etc. W3C insist on plain- test encoding. But people tend to specify XML representation, etc. W3C insist on plain- test encoding. 29

30 2 nd Day Purposes of ontology, taxonomy, data model, etc.? Purposes of ontology, taxonomy, data model, etc.? CAOS vs. TDG6 Sem. Roles? CAOS vs. TDG6 Sem. Roles? A concept is a conjunction of characteristics. … partial equivalence A concept is a conjunction of characteristics. … partial equivalence DC schema and ontology are different. DC schema and ontology are different. But DC metamodel may be defined as an ontology. But DC metamodel may be defined as an ontology. 30

31 Resolutions TDG6 appoints Paul Buitelaar to launch and lead the project for ontologizing LMF, collaborating with Nicoletta Calzorali and other experts. TDG6 appoints Paul Buitelaar to launch and lead the project for ontologizing LMF, collaborating with Nicoletta Calzorali and other experts. 31


Download ppt "ISO TC37/SC4 N429 ISO/TC37/SC4/TDG6 Language Resource Ontologies 2007-11-11/12, Busan 2007-11-11/12, Busan HASIDA Koiti HASIDA Koiti"

Similar presentations


Ads by Google