W3C home > Mailing lists > Public > public-multilingualweb-lt@w3.org > June 2012

ISSUE-22: Provenance, and the W3C PROV-XG

From: Maxime Lefrançois <maxime.lefrancois@inria.fr>
Date: Fri, 8 Jun 2012 20:33:47 +0200 (CEST)
To: David Lewis <dave.lewis@cs.tcd.ie>, Multilingual Web LT Public List <public-multilingualweb-lt@w3.org>
Message-ID: <494384740.4086426.1339180426992.JavaMail.root@inria.fr>
Dear David, all, 

Work session about Provenance will be quite short on Wednesday, so I want to highlight the fact that we could reuse some of the concepts and relations introduced in the W3C provenance working group, at least to define a RDFS vocabulary for ITS 2.0:
http://www.w3.org/TR/prov-dm/#relations-at-a-glance 

Something quite interesting about this group is how they managed to deal with both XML and RDF technologies: 
They built the PROV Data Model http://www.w3.org/TR/2012/WD-prov-dm-20120503/ and then propose : 
PROV-XML, an XML schema for the PROV data model 
PROV-O, the PROV ontology, an OWL-RL ontology allowing the mapping of PROV to RDF
(+ other specifications, there is a list in the primer document http://www.w3.org/TR/prov-primer/)
This is part of what I'll speak about Tuesday morning.


In the PROV Data Model, Different Activities are led by different Agents on the Entities (text fragments, or the document itself). We could a) keep every trace of those activities, or b) keep only the last activity led on the annotated text fragment.

Entities:
ITSEntity ; Document ; Span ; Element ; Text Fragments, ...

Agents: http://www.w3.org/TR/prov-dm/#term-Agent
Three common types of agents are introduced: prov:SoftwareAgent, prov:Organization, prov:Person.
We could introduce typical agents such as: Content Author, Translation Agent, Machine Translation Service, LSP-based Translation Reviewer, Client-based Translation Reviewer, ...
We could also state in the recommendation: "It is acknowledged that these types do not cover all kinds of agent"

Activities:
Quality assurance process, etc.

Thus it would be easy to classify provenance and quality data categories relations
contentLicencingTerm < hasAnnotation (annotation of an entity)
revisionAgent < wasAttributedTo
sourceLanguage < hasAnnotation
translationAgent < wasAttributedTo
qualityError < hasAnnotation (annotation of an activity)
qualityProfile < hasAnnotation (annotation of an entity)
mtConfidence < hasAnnotation (annotation of an activity)
specialRequirement < hasAnnotation (annotation of an entity)
author < hasAnnotation (annotation of an entity) 

about author, it could indeed be considered equivalent to dc:creator : "The person or organization primarily responsible for creating the intellectual content of the resource. For example, authors in the case of written documents, artists, photographers, or illustrators in the case of visual resources." http://dublincore.org/documents/1998/09/dces/


Kind regards,
Maxime Lefrançois 
Ph.D. Student, INRIA - WIMMICS Team 
http://maxime-lefrancois.info 
@Max_Lefrancois 
Received on Friday, 8 June 2012 18:34:16 UTC

This archive was generated by hypermail 2.3.1 : Sunday, 9 June 2013 00:24:56 UTC