Organic.Edunet Repository Tools

1

Organic.Edunet Repository Tools

Hannes Ebnerhebner@kth.se

Knowledge Management Research GroupRoyal Institute of Technology (KTH), Sweden

2

Organic.Edunet

A Multilingual Federation of Learning Repositories with Quality Content for the Awareness and Education of European Youth about Organic Agriculture and Agroecology

3

The Big Picture

4

Content Repository - Conceptual Overview

• Abstraction of quad store and binary data• Separation between:

– Entry (Metametadata, ACL, ...)– Resource– Metadata (local, external, extracted)

• Named graphs

5

Organic.Edunet Application Profile

• Partially the vocabularies are used from LRE v3.0– General Coverage– Metametadata Contribution– Educational

• Learning Resource Type

• Intended End User Role

• Context– Relation Kind

• Creative Commons for Rights Description• Annotation Description holds the quality certification• Ontology terms in Classification Entry

– Organic.Edunet Ontology

6

Metadata Annotation

• Using “Annotation Tool” based on the SHAME code library• AJAX component, embeddable in web applications• Automatically generated user interface• Based on “Annotation Profiles”• Reusable graph patterns

– Based on QEL– Editor for generating APs (e.g. “I need dcterms:title,

dcterms:subject, foaf:Person, and lom:Contribution”

• Metadata is edited directly in the graph• Leaves metadata outside the graph pattern untouched

7

8

9

Metadata Annotation 3/3

10

Validation – Resource Review

11

Ontology Terms

• Organic.Edunet Ontology– Created by domain experts with the help of the technical

partners– Used for the semantic search in the main portal– Assigned like tags– Includes predicates

12

Ontology Term Annotation

13

Interoperability

• RESTful interface– {http­verb} {base­uri}/{context}/{kind}/{entry}

– Context & Entry: identifier– Kind: entry|metadata|...– Example: GET http://oe.confolio.org/23/metadata/2

• Support for multiple formats– JSON/JDIL (targeted towards web applications)– LOM/XML– RDF (based on the IEEE LTSC LOM/DCAM mapping)– Easy to hook in additional converters

• As much HTTP as possible, e.g. content negotiation

14

Harvesting / Querying

• OAI-PMH– Supported formats

• LOM

• Dublin Core– Both directions

• SQI– Translated into SPARQL queries– Layer in between SQI and SPARQL to respect the ACL

• Custom harvesters– FAO Capacity Building Portal

15

Linked Data

• “... a recommended best practice for exposing, sharing, and connecting pieces of data, information, and knowledge ...” (Wikipedia)

• Uses the web to interconnect data which wasn't linked previously

• Based on URIs and RDF• Contextualized information• Not necessary to change the backend

– “Triplification” can be done on demand– Important to use URIs

• SCAM contributes to the Linked Data cloud

16

17

Six months later...

18

19

Current Status

• Intense content population phase

• Number of resources– Harvested: ~6.000– Imported: ~1.000– Target: 10.000+ uploaded and enriched resources

• Harvesting from– Intute (~5.500)– FAO Capacity Building Portal (~500)– FAO Corporate Document Repository

20

Summary

• Web-based tool for learning resource annotation• Transparent application of Semantic Web technologies• Interoperability

– Open standards used wherever possible– LOM/XML <-> DCAM– Linked Data

• Well prepared for federations• Generic applicability

– Not restricted to Organic Agriculture

• Seperate reusable components– Annotation Tool: flexible metadata editor– SCAM: resource and metadata management framework– Confolio: e-portfolio web application (in combination with

SCAM)

21

http://www.organic­edunet.eu