Difference between revisions of "ADAPT2 RDF Binding"

From PAWS Lab
Jump to: navigation, search
(Tentative Conventions)
(Tentative Conventions)
 
Line 11: Line 11:
 
* Schema is written in upper case, e.g. [[DC]], vocabulary in lower case, e.g. [[dc]], [[dcterms]]
 
* Schema is written in upper case, e.g. [[DC]], vocabulary in lower case, e.g. [[dc]], [[dcterms]]
 
* A chain of Object/Property statements is abbreviated with the help of &rarr; (<code>&amp;rarr;</code>), e.g. [[rdf#Description|rdf:Description]]&rarr;[[rdf#type|rdf:type]]&rarr;[[rss#channel|rss:channel]]
 
* A chain of Object/Property statements is abbreviated with the help of &rarr; (<code>&amp;rarr;</code>), e.g. [[rdf#Description|rdf:Description]]&rarr;[[rdf#type|rdf:type]]&rarr;[[rss#channel|rss:channel]]
* Domain-Range pair of a Property is prefixed with :: (double colon) and connected by &rArr; (double rignt arrow), e.g. [[lom#annotation|lom:annotation]] :: [[lom#LearningObject|lom:LearningObject]]&rArr;[[lom#Annotation|lom:Annotation]]
+
* Domain-Range pair of a Property is prefixed with :: (double colon) and connected by &rArr; (<code>&amp;rArr;</code>), e.g. [[lom#annotation|lom:annotation]] :: [[lom#LearningObject|lom:LearningObject]]&rArr;[[lom#Annotation|lom:Annotation]]
 
* RDF-serialization of an ER Element is implemented  - <span style='color:white;background-color:green;padding:2px;'>Done</span>
 
* RDF-serialization of an ER Element is implemented  - <span style='color:white;background-color:green;padding:2px;'>Done</span>
 
* RDF-serialization of an ER Element is (arguably) unnecessary - <span style='color:white;background-color:red;padding:2px;'>Not</span>
 
* RDF-serialization of an ER Element is (arguably) unnecessary - <span style='color:white;background-color:red;padding:2px;'>Not</span>

Latest revision as of 20:44, 19 May 2009

Ambox style.png

This page is under construction and requires cleaning

The main purpose of this effort is to map entity-relationship models of applications in ADAPT2 architecture to RDF. If necessary, appropriate comments would be given about models, schemata and vocabularies themselves.

Further work has the following structural units:

  • ER Models - describe individual adaptive applications. Each entity and relationship is supplied with a list of suggested RDF bindings
  • RDF schemata - describe classes and properties with suggested use in describing entities and relationships of applications. Each schema can have several vocabularies identified by prefixes

Tentative Conventions

  • Objects and Entities are shown in boldface
  • Properties and relationships - regular face
  • Schema is written in upper case, e.g. DC, vocabulary in lower case, e.g. dc, dcterms
  • A chain of Object/Property statements is abbreviated with the help of → (&rarr;), e.g. rdf:Descriptionrdf:typerss:channel
  • Domain-Range pair of a Property is prefixed with :: (double colon) and connected by ⇒ (&rArr;), e.g. lom:annotation :: lom:LearningObjectlom:Annotation
  • RDF-serialization of an ER Element is implemented - Done
  • RDF-serialization of an ER Element is (arguably) unnecessary - Not
  • Decision on RDF serialization needs clarification - Clarify

ER Models

  • Knowledge Tree - course portal (15|0|1|33)
  • WebEx - application serving dissected examples (--|--|--|--)
  • PERSEUS - adaptive functionality profider (--|--|--|--)
  • QuizPACK - a package of parametrized quizzes in C (--|--|--|--)
  • CUMULATE - Centralized User Modeling Server (--|--|--|--)

RDF Schemata and Vocabularies

Polisemy AKA Equivalence

Among various RDF schemata and vocabularies there exist overlap in what they actually mean. For example, dc:title, rdf:label, and rss:title all mean the same. We will maintain (our) classes of equivalence as a reference to RDF homonyms.