iso 16642 - a tutorial part 2: representing data categories
DESCRIPTION
ISO 16642 - a tutorial Part 2: Representing data categories. TMF - Terminological Markup Framework Laurent Romary - Laboratoire Loria. Why formalizing DatCats?. Systematizing data category description: Notion of Data Category Registry (DCR) I need a data category: is it there? - PowerPoint PPT PresentationTRANSCRIPT
ISO 16642 - a tutorialPart 2: Representing data
categories
TMF - Terminological Markup Framework
Laurent Romary - Laboratoire Loria
Why formalizing DatCats?
Systematizing data category description:– Notion of Data Category Registry (DCR)
• I need a data category: is it there?– Query by name, definition etc.
Automatizing processes:– Format control of TMLs– Filters from one TML to GMT
Which model for DatCats?
Using XML:– Coherence with TMF principles– Using stylesheet to generate schemas and filters
Using RDF (Resource Description Framework)– Intended format for representing meta-data:
• Description of a DatCat is meta-data with regards TMF
RDF - a quick presentation
Cf. other file
Data Categories
A Formal Description
Data Category Registry
dcsd:DataCategory
rdf:about
Data Category
DCRegistryDCRegistry
DescriptionDescription
VersionNumber
dcsd:VersionNumber
Data Category description
DCDefinition
DCName
Content
dcsd:DCDefinition
dcsd:DCName
dcsd:Content
dcsd:DCIdentifier
dcsd:Level
DCType (S, C)dcsd:DCType
Salt 2000-11-08/SEW
dcsd:DCAdmin
DCComment
dcsd:DCComment
Data Category
Locus
DCAdmin
DCIdentifierDCParent
dcsd:DCParent
DCExample
dcsd:DCExample
Simple and complex DatCats
Complex data categories– shall serve as field identifiers (not names) in databases
and can have content. The datatype for this content shall be declared for each data category and can commonly take the form of different categories of text, defined data types (such as dates), and specified data domains, e.g., picklists comprising standardized permissible instances.
» Example: /Part of Speech/
Simple data categories– shall serve as the content of complex data categories.
» Example: /Noun/, /Verb/, /Adjective/ etc.
Levels and content
Content
DataType TargetType
Ref to other datcat(s)
dcsd:DataType dcsd:TargetType
rdf:Alt
rdf:li
List of References
List of References
Ref to other datcats
rdf:Alt
rdf:li
Level/Loci
rdf:Alt
Ref to other datcat(s)
rdf:li
List of References
Administrative properties
dcsd:DCAdmin
Data Category
DCAdmin
Status
dcsd:Status
StatusDatedcsd:StatusDate
StatusNote
dcsd:StatusNote
EditionDate
dcsd:EditionDate
ShortForm AdmittedName ForbiddenName
Source
dcsd:Source
VariantNames
dcsd:VariantNames
Dcsd:ShortFormDcsd:AdmittedName Dcsd:ForbiddenName
RDF Representation
/term/ - RDF description (1)
<dcsd:DataCategory dcsd:DCIdentifier="ISO12620A01"dcsd:DCName="term"dcsd:position="A.01"dcsd:DCType="C">
<dcsd:DCDefinition> A verbal designation of a generalconcept in a specific subject field </dcsd:DCDefinition>
<dcsd:DCComment><dcsd:sourceComment>For definition of related term, see
ISO 1087-1, 3.4.3.</dcsd:sourceComment><dcsd:conceptComment>Terms can consist of single words
or be composed of multiword strings…</dcsd:conceptComment><dcsd:Example>"radix" in annex C, figure
C.1.</dcsd:Example><dcsd:DictionnaryID>A.1</dcsd:DictionnaryID>
</dcsd:DCComment>
/term/ - RDF description (2)
<dcsd:Content dcsd:DataType="plainText"/> <dcsd:Level>
<rdf:Alt><rdf:li>TL</rdf:li><rdf:li>TC</rdf:li>
</rdf:Alt></dcsd:Level><dcsd:DCAdmin dcsd:OrgSource="ISO TC 37"
dcsd:DocSource="ISO12620:1999"dcsd:subDate="2000-10-20 SEW"dcsd:registryComment="Prepared
2000-10-20"dcsd:Status="Accepted"/>
</dcsd:DataCategory>
/term type/ - RDF description (1)
<dcsd:DataCategory dcsd:DCIdentifier="ISO12620A0201"dcsd:DCName="term type"dcsd:position="A.02.01"dcsd:DCType="C">
<dcsd:DCDefinition>An attribute assigned to aterm</dcsd:DCDefinition>
<dcsd:DCComment><dcsd:DictionnaryID>A.2.1</dcsd:DictionnaryID>
</dcsd:DCComment><dcsd:Content dcsd:DataType="picklist">
<rdf:Alt><rdf:li>ISO12620A020101</rdf:li><rdf:li>ISO12620A020102</rdf:li><rdf:li>ISO12620A020119</rdf:li>
</rdf:Alt></dcsd:Content>
/term type/ - RDF description (2)
<dcsd:Level><rdf:Alt>
<rdf:li>TL</rdf:li><rdf:li>TC</rdf:li>
</rdf:Alt></dcsd:Level><dcsd:DCAdmin dcsd:OrgSource="ISO TC 37"
dcsd:DocSource="ISO12620:1999"dcsd:subDate="2000-10-20 SEW"dcsd:registryComment="Prepared
2000-10-20"dcsd:Status="Accepted"/>
</dcsd:DataCategory>
Actualizing a DatCat
TMF specific properties
Styling properties
dcsd:Style
Data Category
Style
StyleName
dcsd:StyleName
ElementNamedcsd:ElementName
AttributeName
dcsd:AttributeName
TypeValue
dcsd:TypeValue
Simple
ElementAttribute
TypedElementValuedElementTVElement
Value
dcsd:Value
For ‘ Simple ’
AnchorInfo
dcsd:Anchor
AnchorLevel
Attribute style description
• dcsd:StyleName="Attribute"
– Conditions of use:• Not valid for annotations
– Required properties• dcsd:AttributeName
– Example:• dcsd:AttributeName="id"
• <anchorElement id="xx54893">…</>
Element style description
• dcsd:StyleName="Element"
– Required properties• dcsd:ElementName
– Example:• dcsd: ElementName ="definition"
• <definition>…</definition>
TypedElement style description
• dcsd:StyleName="TypedElement"
– Required properties• dcsd:ElementName, dcsd:TypeValue
– Example:• dcsd:ElementName ="termNote"
• dcsd:TypeValue="partOfSpeech"
• <termNote type="partOfSpeech"/>N</termNote>
ValuedElement style description
• dcsd:StyleName="ValuedElement"
– Conditions of use:• Not valid for annotations
– Required properties• dcsd:ElementName
– Example:• dcsd:ElementName ="pos"
• <pos value="noun"/>
TVElement style description
• dcsd:StyleName="TVElement"
– Conditions of use:• Not valid for annotations
– Required properties• dcsd:ElementName, dcsd:TypeValue
– Example:• dcsd:ElementName ="free"• dcsd:TypeValue="pos"
• <free type="pos" value="noun"/>
Simple style description
• dcsd:StyleName="Simple"
– Conditions of use:• Express the value of simple data categories
– Required properties:• dcsd:Value
– Example:• dcsd:Value ="Nom"
• <pos>Nom</pos>
Dealing with languages
Two types of languages
Working language• The language used at a given place in a document,
along the XML hierarchy
• Representation: xml:lang
Object language• The language about which you speak at a given place
in your terminological entry (e.g. describes the Language Section level)
• Representation: as a data category "language", with a narrow scope
Example — DXLT
<langSet lang='en’ xml:lang="fr"><descrip type="definition">Une valeur entre 0 et 1 utilisée...</descrip><tig>
<term xml:lang="en">alpha smoothing factor</term>
<termNote type="termType">fullForm</termNote></tig>
</langSet>
Example — GMT
<struct type="LS" xml:lang="fr"><feat type="language">en</feat><feat type="definition">Une valeur entre 0 et 1 utilisée...</feat><struct type="TL">
<feat type="term" xml:lang="en">alpha smoothing factor</feat>
<feat type="termType">fullForm</feat></struct>
</langSet>
Conclusion
– A general model for analysing and representing terminological data collection
– An underlying formalism expressed in XML,RDF
– Associated tools (Salt project)• DCSEditor,
• DCSBrowser,
• Automatic generation of XSLT filters and XML schemas from a given TML specification
Useful pointers
SALT project– http://www.loria.fr/projets/SALT– http://www.ttt.org/
The TMF site– http://www.loria.fr/projets/TMF