Skip to content

Commit

Permalink
Notes pointing out added property uses
Browse files Browse the repository at this point in the history
  • Loading branch information
Simon Cox committed Jul 5, 2018
1 parent c0a4cab commit 1bccc16
Showing 1 changed file with 49 additions and 32 deletions.
81 changes: 49 additions & 32 deletions dcat/index.html
View file
Original file line number Diff line number Diff line change
Expand Up @@ -979,6 +979,11 @@ <h4>Property: dataset relation</h4>
Since <a href="http://purl.org/dc/terms/relation">dct:relation</a> is a super-property of <a href="http://purl.org/dc/terms/conformsTo">dct:conformsTo</a>, <a href="http://www.w3.org/ns/dcat#distribution">dcat:distribution</a>, <a href="http://purl.org/dc/terms/hasPart">dct:hasPart</a>, <a href="http://purl.org/dc/terms/references">dct:references</a> and <a href="http://purl.org/dc/terms/requires">dct:requires</a>,
it may be used to link a dataset to various component and supplementary resources, without being inconsistent with a subsequent reclassification to add more specific semantics.
</p>

<p class="note">
Use of this Dublin Core Terms property in this context added in this revision of DCAT.
</p>

</section>

<section id="Property:dataset_distribution">
Expand Down Expand Up @@ -1011,6 +1016,10 @@ <h4>Property: dataset part</h4>
</tbody>
</table>

<p class="note">
Use of this Dublin Core Terms property in this context added in this revision of DCAT.
</p>

</section>

<section id="Property:dataset_references">
Expand All @@ -1024,6 +1033,10 @@ <h4>Property: dataset reference</h4>
</tbody>
</table>

<p class="note">
Use of this Dublin Core Terms property in this context added in this revision of DCAT.
</p>

</section>

<section id="Property:dataset_requires">
Expand All @@ -1037,6 +1050,10 @@ <h4>Property: dataset requirement</h4>
</tbody>
</table>

<p class="note">
Use of this Dublin Core Terms property in this context added in this revision of DCAT.
</p>

</section>

<section id="Property:dataset_frequency">
Expand Down Expand Up @@ -1716,23 +1733,23 @@ <h2>Quality information</h2
<div class="note">
<p>This section is not-normative as it provides guidance on how to document the quality of DCAT first class entities (e.g., datasets, distributions) and it does not define new DCAT terms. The guidance relies on the Data Quality Vocabulary(DQV)[[vocab-dqv]], which is a W3C Group Note.</p>
</div>

<p class="issue" data-number="58">
The need to choose or define a data <b>quality model</b> has been identified as a requirement to be satisfied in the revision of DCAT.
</p>
</p>
The Data Quality Vocabulary (DQV) offers common modelling patterns for different aspects of Data Quality.

It can relate DCAT datasets and distributions with different types of quality information including
<ul>
<li> <a href="https://www.w3.org/TR/vocab-dqv/#dqv:QualityAnnotation"> dqv:QualityAnnotation</a>, which represents feedback and quality certificates given about the dataset or its distribution. </li>
<li> <a href="https://www.w3.org/TR/vocab-dqv/#dqv:QualityAnnotation"> dqv:QualityAnnotation</a>, which represents feedback and quality certificates given about the dataset or its distribution. </li>
<li> <a href="https://www.w3.org/TR/vocab-dqv/#dqv:QualityPolicy">dqv:QualityPolicy</a>, which represents a policy or agreement that is chiefly governed by data quality concerns.</li>
<li><a href="https://www.w3.org/TR/vocab-dqv/#dqv:QualityMeasurement">dqv:QualityMeasurement</a>, which represents a metric value providing quantitative or qualitative information about the dataset or distribution.</li>
</ul>

Each type of quality information can pertain to one or more quality dimensions, namely, quality characteristics relevant to the consumer. The practice to see the quality as a multi-dimensional space is consolidated in the field of quality management to split the quality management into addressable chunks. DQV does not define a normative list of quality dimensions. It offers the quality dimensions proposed in ISO/IEC 25012 [[ISOIEC25012]] and Zaveri et al. [[ZaveriEtAl]] as two possible starting points. It also provides an <a href="https://www.w3.org/2016/05/ldqd">RDF representation</a> for the quality dimensions and categories defined in the latter. Ultimately, implementers will need to choose themselves the collection of quality dimensions that best fits their needs.
Each type of quality information can pertain to one or more quality dimensions, namely, quality characteristics relevant to the consumer. The practice to see the quality as a multi-dimensional space is consolidated in the field of quality management to split the quality management into addressable chunks. DQV does not define a normative list of quality dimensions. It offers the quality dimensions proposed in ISO/IEC 25012 [[ISOIEC25012]] and Zaveri et al. [[ZaveriEtAl]] as two possible starting points. It also provides an <a href="https://www.w3.org/2016/05/ldqd">RDF representation</a> for the quality dimensions and categories defined in the latter. Ultimately, implementers will need to choose themselves the collection of quality dimensions that best fits their needs.

The following section shows how DCAT and DQV can be coupled to describe the quality of datasets and distributions.
For a comprehensive introduction and further examples of use, please refer to the Data Quality Vocabulary (DQV) group note [[vocab-dqv]].
The following section shows how DCAT and DQV can be coupled to describe the quality of datasets and distributions.
For a comprehensive introduction and further examples of use, please refer to the Data Quality Vocabulary (DQV) group note [[vocab-dqv]].
<div class="note">
<p>The following examples make no comments on where the quality information would reside and how it is managed. That is out of scope for the DCAT vocabulary. The assumption made is that the quality individuals are available using the URIs indicated.
Besides, the examples and more in general the DQV is neutral to the data portal design choices on how to collect quality information. For example, data portals can collect DQV instances by implementing specific UI to annotate data or by taking inputs from 3rd-party services.
Expand All @@ -1741,19 +1758,19 @@ <h2>Quality information</h2

<p class="issue" data-number="252">
We might want to include examples of quality documentation related to services.
</p>
</p>

<section id="quality-example1">
<h2>Providing quality information</h2>
<p class="issue" data-number="57">
The need to provide hook for <b>quality</b> information concerning a dcat:Dataset has been identified as a requirement to be satisfied in the revision of DCAT.
</p>
A data consumer (:consumer1) describes the quality of the dataset :genoaBusStopsDataset that includes a georeferenced list of bus stops in Genoa. He/she annotates the dataset with a DQV quality note (:genoaBusStopsDatasetCompletenessNote) about data completeness (ldqd:completeness) to warn that the dataset includes only 20500 out of the 30000 stops.

<pre>:genoaBusStopsDataset a dcat:Dataset ;
dqv:hasQualityAnnotation :genoaBusStopsDatasetCompletenessNote .

:genoaBusStopsDatasetCompletenessNote
:genoaBusStopsDatasetCompletenessNote
a dqv:UserQualityFeedback ;
oa:hasTarget :genoaBusStopsDataset ;
oa:hasBody :textBody ;
Expand All @@ -1765,9 +1782,9 @@ <h2>Providing quality information</h2>

:textBody a oa:TextualBody ;
rdf:value "Incomplete dataset: it contains only 20500 out of 30000 existing bus stops" ;
dc:language "en" ;
dc:format "text/plain"
.
dc:language "en" ;
dc:format "text/plain"
.
</pre>
The activity :myQualityChecking employs the service :myQualityChecker to check the quality of the :genoaBusStopsDataset dataset. The metric :completenessWRTExpectedNumberOfEntities is applied to measure the dataset completeness (ldqd:completeness) and it results in the quality measurement :genoaBusStopsDatasetCompletenessMeasurement.
<pre>:genoaBusStopsDataset
Expand All @@ -1780,18 +1797,18 @@ <h2>Providing quality information</h2>
dqv:value "0.6833333"^^xsd:decimal ;
prov:wasAttributedTo :myQualityChecker ;
prov:generatedAtTime "2018-05-27T02:52:02Z"^^xsd:dateTime ;
prov:wasGeneratedBy :myQualityChecking
prov:wasGeneratedBy :myQualityChecking
.

:completenessWRTExpectedNumberOfEntities
:completenessWRTExpectedNumberOfEntities
a dqv:Metric ;
skos:definition "it returns the degree of completeness as ratio between the actual number of entities included in the dataset and the declared expected number of entities."@en ;
dqv:expectedDataType xsd:decimal ;
dqv:inDimension ldqd:completeness .

# :myQualityChecker is a service computing some quality metrics
# :myQualityChecker is a service computing some quality metrics
:myQualityChecker
a prov:SoftwareAgent ;
a prov:SoftwareAgent ;
rdfs:label "A quality assessment service"^^xsd:string .
# Further details about quality service/software can be provided, for example,
# deploying vocabularies such as Dataset Usage Vocabulary (DUV), Dublin Core or ADMS.SW
Expand All @@ -1806,7 +1823,7 @@ <h2>Providing quality information</h2>
prov:endedAtTime "2018-05-27T02:52:02Z"^^xsd:dateTime;
prov:startedAtTime "2018-05-27T00:52:02Z"^^xsd:dateTime .
</pre>

</section>
</section>
<section id="prov-patterns">
Expand Down Expand Up @@ -1921,10 +1938,10 @@ <h2>Relation to other W3C Recommendations</h2>
<p class="issue" data-number="254">
DCAT should be aligned with other recent Linked Data based Recommendations.
</p>

<section id="ldp">
<h3>Linked Data Platform (LDP)</h3>

<p>
DCAT provides a data model for representation of metadata about datasets in the form of Linked Data, but it does not specify how this metadata should be accessed or modified.
The DCAT compatible metadata can be viewed as collections of Catalog Records, Datasets and Data Services contained in a Catalog, and a collection of Distributions contained in a Dataset.
Expand All @@ -1951,15 +1968,15 @@ <h3>Linked Data Platform (LDP)</h3>
&lt;&gt; a dcat:Catalog ;
dcat:datasets &lt;/datasets/&gt; ;
dcat:dataset &lt;/datasets/001&gt; .

&lt;/datasets/&gt; a ldp:Container, ldp:DirectContainer ;
ldp:membershipResource &lt;&gt; ;
ldp:hasMemberRelation dcat:dataset ;
ldp:contains &lt;/datasets/001&gt; .

&lt;/datasets/001&gt; a dcat:Dataset .</pre>
</div>

<p>
In the second example, we add LDPCs <code>&lt;/records/&gt;</code> for Catalog Records and <code>&lt;/services/&gt;</code> for Data Services, discoverable using <code>dcat:records</code> and <code>dcat:services</code> predicates from the Catalog:
</p>
Expand All @@ -1978,7 +1995,7 @@ <h3>Linked Data Platform (LDP)</h3>
dcat:datasets &lt;/datasets/&gt; ;
dcat:services &lt;/services/&gt; ;
dcat:dataset &lt;/datasets/001&gt; .

&lt;/records/&gt; a ldp:Container, ldp:DirectContainer ;
ldp:membershipResource &lt;&gt; ;
ldp:hasMemberRelation dcat:record ;
Expand All @@ -1988,15 +2005,15 @@ <h3>Linked Data Platform (LDP)</h3>
ldp:membershipResource &lt;&gt; ;
ldp:hasMemberRelation dcat:dataset ;
ldp:contains &lt;/datasets/001&gt; .

&lt;/services/&gt; a ldp:Container, ldp:DirectContainer ;
ldp:membershipResource &lt;&gt; ;
ldp:hasMemberRelation dcat:service ;
ldp:contains &lt;/services/001&gt; .

&lt;/records/001&gt; a dcat:CatalogRecord ;
foaf:primaryTopic &lt;/datasets/001&gt; .

&lt;/datasets/001&gt; a dcat:Dataset ;

&lt;/services/001&gt; a dcat:DataService .</pre>
Expand Down Expand Up @@ -2026,14 +2043,14 @@ <h3>Linked Data Platform (LDP)</h3>

&lt;/datasets/001/distributions/001&gt; a dcat:Distribution .</pre>
</div>
<p class="note">For catalogs with many datasets, catalog records, data services or distributions,

<p class="note">For catalogs with many datasets, catalog records, data services or distributions,
the Linked Data Platform Paging mechanism [[ldp-paging]] SHOULD be used to provide access to them.</p>

<p>
In the next sections we formally define the additional properties used for discovery of LDP containers.
</p>

<section id="Property:catalog_datasets">
<h4>Property: datasets</h4>

Expand Down Expand Up @@ -2090,10 +2107,10 @@ <h4>Property: distributions</h4>

<section id="ldn">
<h3>Linked Data Notifications (LDN)</h3>

<p>
Linked Data Notifications (LDN) [[ldn]] can be used with DCAT e.g. for feedback collection.
Any resource can have an LDN Inbox.
Any resource can have an LDN Inbox.
In the following example we show a dataset <code>&lt;/datasets/001&gt;</code> as an LDN Target with an LDN Inbox.
</p>
<div class="example">
Expand Down

0 comments on commit 1bccc16

Please sign in to comment.