FAqT Vocabulary

What's first

• Getting started

What we'll cover

Let's get to it

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. ?

What's next