Data quality draft actions

Actions for the DQV draft, not tracked by formal actions

NB: more formal actions are tracked at https://www.w3.org/2013/dwbp/wiki/Data_quality_schedule

• [done] remove Makx' affiliation
• [done] Typo: in 6.1 the example listing has :cvsAvailabilityMetric which I think should be :csvAvailabilityMetric
• [done] reflect the conclusion of http://www.w3.org/2013/dwbp/track/actions/188
  • [done] section 3, issue 2 can be removed assumed we make sure we have provided the proper credits to DAQ, in this respect, daq papers has been added as references.
  • [done] section 4 is not are not yet consistent, Fig 1 should be replaced with the in progress new figure [1]
  • [done] section 5 has been updated but we need to doublecheck inverse properties, subproperties and subclasses
  • [done] section 6: examples should be updated turning daq into dqv )
• [done] remove link to the issue tracker on github
• [done] updated draft considering decision on issue 182 and 186
• [done] An issue should be raised, asking in the doc whether dqv:QualityPolicy should be a sub-class of dcterms:Standard
• [done] to be able to close ISSUE-185 (and remove the corresponding note!)
  • [done] dqv:QualityAnnotation is a subclass of oa:Annotation. The instances of this class should have one oa:motivatedBy statement with a an instance of oa:Motivation (and skos:Concept), which reflects a quality assessment purpose. We define this instance as dqv:qualityAssessment
  • [done] dqv:QualityAnnotation should be formally (i.e in OWL) defined as equivalent to the class of all oa:Annotation that have an oa:motivatedBy statement set to dqv:qualityAssessment.
• [done] following closure of ISSUE-165: rename DQV:UserFeedback with dqv:QualityUserFeedback making it as duv:Feedback subclass [updated considering also the schema |F2F integration between duv and dqv
• [done] turned the issue box for Issue-179 into a editorial note, considering the resolution 8
• [done] following the closure of issue-187, and resolution 9: Don't keep the constraints from DAQ but provide guidelines
  • [commented constraints in the html, in case we want to use them later] we are expected to delete the cardinality constraints
  • [done] to put a note to provide some guidance about property cardinalities
• [done] Write a sentence in the requirement section that said that we explored other use cases, but this was not a mandatory. And remove the note about Issue-190
• [done] Remove the note in the namespace section on Issue 179
• [done] to add an annex showing the mapping between zaveri and ISO dimensions
• [done] replace QualityMeasure by QualityMeasurement. and dqv:QualityMeasureDataset by dqv:QualityMeasurementDataset, dqv:hasQualityMeasure by dqv:hasQualityMeasurement [done]
• [done] Revise example involving qualityMeasurements, dimension and metrics in order to make them more informative ( e.g. by adding proper skos:definition skos:prefLabel)
• [done] replace current dqv:ServiceLevelAgreement with dqv:QualityPolicy (http://www.w3.org/2013/dwbp/track/issues/184)
• [done] Add an example with an SLA as Quality Policy, trying to use the same dimensions as metrics and annotations resolution 3 and action 203
• [done] comment all the part mentioning abstract classes / properties in the document, including *class/ property descriptions in the linkset example
• [done] add example for section 5.3 Document the provenance of single quality measurement
• [done] Fix text following Comment from Giancarlo Guizzardi: A Metric is not a unit of measurement. An Observation (QualityMeasure) assigns a value in a given unit to a Metric.
  • [done] AI: actually the definitions of Metric and QualityMeasures should be revised. Currently the definition of QualityMeasure says more about metrics than the definition of Metric!
  • [done] This may need to be adapted to the change of QualityMeasure into QualityMeasurement.
  • [done] implement proposal/resolution on issue 231 [ https://lists.w3.org/Archives/Public/public-dwbp-wg/2016Feb/0070.html ]
• [done] Re-draft section 6
  • [done] to draft a new version of section 6 adding zaveri and iso dimensions, add example and fix typos
  • [done] to prepare ttl for zaveri and Iso dimensions
    • [done] created a ttl for zaveri dimension linked to the DQV document
    • [done] to decide in which (stable) namespace we can put zaveri.ttl -> RI: we decided for https://www.w3.org/2016/05/ldqd
• [done] add property dqv:expectedDataType in metric and example
  • [done - no change ] it is equivalent to daq:expectedDataType depending on how we will solve Issue-224.
• [done] revise example on linkset quality
  • [it does not apply in the current example] use dcterms:LinguisticSystem not madsrdf:Language
  • [done] implement now your suggestion in the email on the use 'onLanguage' and 'onProperty' as property names, instead of the very generic-sounding 'language' and 'property'.
• [done] add an example use SKOS semantic relationship to indicate specialization links inside any of the three levels, as we've penciled in a note at http://w3c.github.io/dwbp/vocab-dqg.html#DimensionsOfISOIEC25012
  • maybe add a paragraph about this in the section about dimension and categories.
  • AI: I think this is not needed anymore
• [done] Ask to zaveri et al lf they are happy with our RDF representation and examples, and want to suggest changes
• [done] move Guéret in the list of contributors
• [done] add an acknowledgment to all the group members and people who have sent feedback (i.e., Andrea Perego, Werner Bailer, Guillaume Duffes...)
• [done] update history of changes
• [done] In classes Metric, Dimension, Category, we should consider to add also properties skos:definition and skos:prefLabel ( Issue 204 and 205 are still pending... but these properties are already used in the examples.) or a sub-section about using SKOS in general?
  • [done] Antoine should write a specific section about using SKOS for documentation and linking, a bit longer and visible than a usage note
• [done] consider if changing wording of labels and choice of instance names (which still 'look' generic, i.e. they could be names for classes) for examples 5.3 and 5.2.
• [done] to split the example about quality linkset in two sub sections, one for parameters and one for representation in RDFcube
• [done] introduce a table for dqv:QualityPolcy
• [done] introduce a table for dqv:QualityCertificate
• [done] following closure of issue-164, issue-189, and resolution 10: keep dqv as it is, provide guidance on how daq can work with another quality statistics vocabulary , .... revise the paragraph 7.1 Statistics:
  • [done] saying: although statistics are not actual quality measures they bring useful input for quality evaluation, in the quality jargon they can be seen as "quality indicator"
  • [done] adding a note saying " once the DQV is defined, we are planning to provide guidance on how dqv can work with another quality statistics vocabulary "
  • [done] deciding which quality statistics vocabulary we want to consider [ e.g., aether, lodstats??? we need an input from the group i guess - AI: I think we should keep aether as it was the one motivating issue-189]
  • [done] Annette had suggested to "include some statistics that are more directly related to quality than the examples, which seem text-centric. Percentage of empty values and percentage of nulls come to mind."
• [done] add a table defining dqv:hasQualityAnnotation and axiom SubObjectPropertyOf( dqv:hasQualityAnnotation ObjectInverseOf( oa:hasTarget ) )
• [done] Add a note about representing provenance of quality measurements even if a class is not a sub-class of prov:Entity, as per Riccardo's answer to Jeremy
• [done] Consider to add an explicit link to DQV implementation wiki page [2] as per Riccardo's answer to Jeremy
• [done] revise 5.6 Express a question about dataset quality--> in the latest w3c working note https://www.w3.org/TR/annotation-vocab/ they have a new class https://www.w3.org/TR/annotation-vocab/#textualbody, we might consider. [AI: oh yes we should remove cnt:ContentAsText!]
• [done] About section 4,
  • [done] Should we also add descriptions for prov:wasAttributeTo, prov:wasGeneratedBy, qb:structure, odrl:target ?
  • [done] Delete note at the beginning of section "This section is work in progress. We will include later more tables with specification of individual classes and properties."
  • [done] Remove issue 1, about issue 223 and parameter, we have an appendix discussing this issue -> AI: I think we should keep a note to this issue as a postponed one, or a new (sub)issue cf discussion on mailing list.--> RA: I have deleted it as this issues is still mentioned in the Appendix "Defining and using parameters for metrics" and it is mentioned is the wish list at [3].
  • [done] for each class added a sentence introducing the properties that can be used
  • [done] for each class or property having no more than a couple of example in the document ( see association between classes/properties and examples) added explicit reference to the section providing the example.
  • [done] RA: we shoudl decide if change the definition of dqv:QualityMeasurementDataset from "Represents a dataset of quality measurements, evaluations of a given dataset (or dataset distribution) against a specific quality metric" to "Represents a dataset of quality measurements, evaluations of one or more datasets (or dataset distributions) against specific quality metrics".
• [done] rename "Usage Note" into "DQV Usage Note"?
• [done] Add a proper skos:ConceptScheme for dimensions and categories defined in https://www.w3.org/2016/05/ldqd
• [done] Add Vladimir Alexiev in the acknowledgments.
• [done - no change] consider to have two distinct arrows for the oa:hasTarget and dqv:hasQualityAnnotation in the DQV diagram. As they are now, it seems that they are one the inverse of the other, while the latter is a subproperty of the inverse of the former.
• [done] re-think of how to enhance wording - and perhaps ordering - of the content of appendix D on parameters, following Werner's comments (see https://lists.w3.org/Archives/Public/public-dwbp-wg/2016Jul/0013.html)
• [Done] as part of the sub-section on using SKOS to document dimensions, add something about using Dublin Core on QualityMeasurements https://lists.w3.org/Archives/Public/public-dwbp-comments/2016Aug/0029.html
• [Done] update the sentence "The model for the Data Quality Vocabulary is nearing maturity, but the Working Group is seeking feedback on a number of specific issues highlighted in the document below." which appears in section "Status of This Document".
• [Done ] delete reference to issue 221
• [Done ] update the section "Change History".
• [Done ] Homogeneize capitalization of property section headings?
• [Done ] revert the replacement of a skos:related link between ldqd:availability and ldqd:interlinking by one between ldqd:availability and ldqd:semanticAccuracy. ?
• [Done ] Check what's expected for the de-referecing behaviour of the ldf.fi URIs in the stats section.
• [Dropped] Add a dqv:computedOn arrow?
• [Done ] consider to add the clarifications asked by Vladimir Alexiev, " about the difference between the QualityMeasurement and QualityAnnotation, and emphasize that QualityMeasurement is used with literal values, while QualityAnnotation is typically used with resource values. It would also be great to give an example with using nominal (resource) values for clarification."
• [done] why the colours that appears in the ttl examples are so inconsistent?
• [moved to wishlist] discuss adding attributes for the severity of a quality problem, as per discussion with Amrapali Zaveri https://lists.w3.org/Archives/Public/public-dwbp-comments/2016Aug/0028.html
• [moved to wishlist] discuss adding attributes for the 'provenance' of a quality measurement in a part of a dataset, as per discussion with Amrapali Zaveri https://lists.w3.org/Archives/Public/public-dwbp-comments/2016Aug/0028.html
• [dropped] add a reference to Bruce, Thomas R. Metadata quality in a linked data Context. VOXPOPULII Blog (Cornell Legal Information Institute. Jan. 24, 2013.) ?

Riccardo's questions

• [moved to wishlist] should we rename QualityCertificate? the current name is a little misleading, it seems it is a quality certificate rather then an annotation pointing to a quality certificate .

Antoine's questions

• [done] AI has asked Amrapali and Anisa some questions about the mapping table between their dimensions and ISO

update dqv.ttl

• [done] replace QualityMeasure by QualityMeasurement. and dqv:QualityMeasureDataset by dqv:QualityMeasurementDataset, dqv:hasQualityMeasure by dqv:hasQualityMeasurement
• [done] implement proposal/resolution on issue 231 [ https://lists.w3.org/Archives/Public/public-dwbp-wg/2016Feb/0070.html ]
• [done] add property dqv:expectedDataType in metric
• [done] introduce the class dqv:QualityCertificate
• [done] introduce the class dqv:QualityPolcy
• [done] add dqv:hasQualityAnnotation and axiom SubObjectPropertyOf( dqv:hasQualityAnnotation ObjectInverseOf( oa:hasTarget ) )
• [done] add the axiom SubObjectPropertyOf( ObjectInverseOf( daq:hasMetric )  :inDimension)
• [done] add dqv:hasQualityMetadata
• [done] update Turtle file with changes in the classes/property tables (e.g. DQV definition).
• [dropped] add the vocabulary html descriptions to be returned by HTTP negotiation
• [moved to wishlist] add multilingual translations as soon as we finish DQV translations

update example in BP

• [done] replace QualityMeasure by QualityMeasurement. and dqv:QualityMeasureDataset by dqv:QualityMeasurementDataset, dqv:hasQualityMeasure by dqv:hasQualityMeasurement
• [done] implement proposal/resolution on issue 231 [ https://lists.w3.org/Archives/Public/public-dwbp-wg/2016Feb/0070.html ]