FAqT Vocabulary

How will the dataset evaluation service describe its result? The following models overlap in different ways, so the answer to what at FAqT should return is probably a coherent mix of:

• Data Quality Ontology (DQ)
• RDF Data Cube
• Measurement Unit Ontology
• Linked Open (LODQ)
• subjective opinions expressed on the web (MARL)
• http://labs.mondeca.com/sparqlEndpointsStatus/endpoint/endpoint.html
• Bernard Vatant: DQM vocabulary and also to the LOV search engine.
• http://twitter.com/RubenVerborgh looking for parameters (availability, performance, easy-of-use) to describe the server or actor that generates result sets. any clues?

(TODO: make diagrams for these)

Data Quality Ontology

Current location in svn, this will likely be moved into its own subdirectory of public/ontologies/SESF/DQ at some point in the near future

https://scm.escience.rpi.edu/svn/public/ontologies/STOM/trunk/dq.owl

The (incomplete) primer. The intuitive description of the core concepts is complete, but subject to updates.

http://tw.rpi.edu/web/project/MDSA/ProductQualityModelPrimer

I'll send some rdf examples by wednesday of next week.

--Stephan

RDF Data Cube

http://publishing-statistical-data.googlecode.com/svn/trunk/specs/src/main/html/cube.html

Measurement Unit Ontology

http://forge.morfeo-project.org/wiki_en/index.php/Units_of_measurement_ontology

Linked Open Data Quality

http://logd.tw.rpi.edu/lodq

MARL

http://marl.gi2mo.org/?page_id=1

Bottom-up requirements

1. Satisfactory vs. Unsatisfactory

example: redirect loop exists or does not

2. Quantitative Ratio with Subjective Assessment (include BOTH so others can reassess the quantitative)

example: percentage of vocab URIs that resolve to RDF descriptions (72% do, service demands 95% for Satisfactory, but a user's query uses vocab within the 95% so Satisfactory).

3. ?