forked from W3C-HCLSIG/HCLSDatasetDescriptions
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathOverview.html
1746 lines (1424 loc) · 96.5 KB
/
Overview.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<?xml version="1.0" encoding="utf-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
<head>
<title>Dataset Descriptions: HCLS Community Profile</title>
<!--
TODO list:
* fix References, citations
* left-align "Description" field of RDF bundle table
-->
<style type="text/css">
table {
border-collapse:collapse;
}
td, th {
border:1px solid black;
padding:1em;
}
table#namespaces td {
font-family: monospace;
}
table.definition {
width:100%;
}
table.definition td.prop {
width:10em;
}
/* .editorsnote::before {
content: "Editor's Note";
display: block;
width: 150px;
background: #F30023;
color: #fff;
margin: -1.5em 0 0.5em 0;
font-weight: bold;
border: 1px solid #cff6d9;
padding: 3px 1em;
}
.editorsnote {
margin: 1em 0em 1em 1em;
padding: 1em;
border: 2px solid #cff6d9;
} */
pre {
padding: 1em;
border: 1px dashed #2f6fab;
color: black;
background-color: #f9f9f9;
line-height: 1.1em;
}
li.tocline1 {margin-left: 2ex}
li.tocline2 {margin-left: 5ex}
.todo {
color: red;
background-color: yellow;
font-weight: bold;
}
</style>
<link rel="stylesheet" type="text/css" href="local.css" />
<!-- link rel="stylesheet" type="text/css" href="http://www.w3.org/StyleSheets/TR/Note" / -->
<link rel="stylesheet" type="text/css" href="http://www.w3.org/StyleSheets/TR/base" />
</head>
<body>
<div class="head">
<p><a href="http://www.w3.org/">
<img src="http://www.w3.org/Icons/w3c_home" alt="W3C" height="48" width="72" /></a></p>
<h1 id="main">Dataset Descriptions: HCLS Community Profile</h1>
<!-- h2 class="no-num no-toc" id="w3c-doctype">W3C Working Draft 4 April 2008</h2 -->
<dl>
<dt>Editors working draft.</dt>
<dt>Editors:</dt>
<dd>Alasdair J.G. Gray, Heriott-Watt University, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Michel Dumontier, Stanford University, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>M. Scott Marshall, MAASTRO Clinic, The Netherlands <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Joachim Baran, Stanford University, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dt>Contributors:</dt>
<dd>Peter Ansell, CSIRO, Australia <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Gary D. Bader, The Donnelly Centre, University of Toronto, Canada <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Asuka Bando, NBDC, Japan <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Alison Callahan, Carleton University, Canada <<a href="[email protected]">[email protected]</a>></dd>
<dd>José Cruz-Toledo, Carleton University, Canada <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Erich Gombocz, IO Informatics, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Alejandra Gonzalez-Beltran, University of Oxford, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Paul Groth, VU University Amsterdam, The Netherlands <<a href="mailto:[email protected]"> [email protected]</a>></dd>
<dd>Melissa Haendel, Oregon Health and Science University, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Maori Ito, NIBIO, Japan <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Simon Jupp, EMBL-EBI, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Toshiaki Katayama, Database Center for Life Sciences, Japan <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Kalpana Krishnaswami, Metaome, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Simon Lin, Marshfield Clinic Research Foundation, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Chris Mungall, Lawrence Berkeley National Laboratory, USA <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Nicolas Le Novere, EMBL-EBI, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Camille Laibe, EMBL-EBI, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Nick Juty, EMBL-EBI, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>James Malone, EMBL-EBI, UK <<a href="mailto:[email protected]">[email protected]</a>></dd>
<dd>Laurens Rietveld, VU University Amsterdam, The Netherlands <<a href="mailto:[email protected]">[email protected]</a>></dd>
</dl>
<p class="copyright"><a href="http://www.w3.org/Consortium/Legal/ipr-notice#Copyright">Copyright</a> © 2014 <a href="http://www.w3.org/"><acronym title="World Wide Web Consortium">W3C</acronym></a><sup>®</sup> (<a href="http://www.csail.mit.edu/"><acronym title="Massachusetts Institute of Technology">MIT</acronym></a>, <a href="http://www.ercim.org/"><acronym title="European Research Consortium for Informatics and Mathematics">ERCIM</acronym></a>, <a href="http://www.keio.ac.jp/">Keio</a>), All Rights Reserved. W3C <a href="http://www.w3.org/Consortium/Legal/ipr-notice#Legal_Disclaimer">liability</a>, <a href="http://www.w3.org/Consortium/Legal/ipr-notice#W3C_Trademarks">trademark</a> and <a href="http://www.w3.org/Consortium/Legal/copyright-documents">document use</a> rules apply.</p>
</div>
<hr title="Separator for header" />
<div>
<h2 class="notoc" id="abstract">Abstract</h2>
<p>Access to consistent, high-quality metadata is critical to finding, understanding, and reusing scientific data. This document describes a consensus among participating stakeholders in health care and the life sciences domain on the description of datasets using the Resource Description Framework (RDF). This specification meets key functional requirements, reuses existing vocabularies to that extent that it is possible, and addresses elements of data description, versioning, provenance, discovery, exchange, query, and retrieval.
</p>
</div>
<div>
<h2 id="status">Status of This Document</h2>
<p>This is an editor's draft of a document for the <a href="../..">Semantic Web in Health Care and Life Sciences Interest Group (HCLSIG)</a>. This is a live document and is subject to change without notice. It reflects the best effort of the editors to reflect implementation experience and incorporate input from various members of the IG, but is not yet endorsed by the IG as a whole. This document is not endorsed by the W3C or its member companies.</p>
</div>
<hr />
<div class="toc">
<h2 id="contents">Table of Contents</h2>
<ul class="toc">
<li class="tocline1"><a href="#introduction">1 Introduction</a> </li>
<li class="tocline1"><a href="#scope">2 Scope</a> </li>
<li class="tocline1"><a href="#conventions">3 Conventions</a> </li>
<li class="tocline1"><a href="#functionalrequirements">4 Functional Requirements</a> </li>
<li class="tocline2"><a href="#s4_1">4.1 Dataset Identification, Description, Licensing, and Provenance</a></li>
<li class="tocline2"><a href="#s4_2">4.2 Dataset Discovery (via Catalog)</a></li>
<li class="tocline2"><a href="#s4_3">4.3 Exchange of Dataset Descriptions</a></li>
<li class="tocline2"><a href="#s4_4">4.4 Dataset Linking</a></li>
<li class="tocline2"><a href="#s4_5">4.5 Content Summary</a></li>
<li class="tocline2"><a href="#s4_6">4.6 Monitoring of Dataset Changes</a></li>
<li class="tocline1"><a href="#datasetdescriptionlevels">5 Dataset Description Levels</a> </li>
<li class="tocline1"><a href="#metadataguidancenotes">6 Metadata Guidance Notes</a> </li>
<li class="tocline2"><a href="#s6_1">6.1 Literals</a></li>
<li class="tocline2"><a href="#s6_2">6.2 Core Metadata</a></li>
<li class="tocline2"><a href="#s6_3">6.3 Identifiers</a></li>
<li class="tocline2"><a href="#s6_4">6.4 Provenance and Change</a></li>
<li class="tocline2"><a href="#s6_5">6.5 Availability and Distributions</a></li>
<li class="tocline2"><a href="#s6_6">6.6 Statistics</a></li>
<li class="tocline1"><a href="#toolingsupport">7 Tooling Support</a> </li>
<li class="tocline2"><a href="#s7_1">7.1 Dataset Description Generation Tool</a></li>
<li class="tocline2"><a href="#s7_2">7.2 Validation Tool</a></li>
<li class="tocline1"><a href="#usecases">8 Use Cases</a> </li>
<li class="tocline2"><a href="#s8_1">8.1 Marshfield Use Case</a></li>
<li class="tocline2"><a href="#s8_2">8.2 Metaome Transcriptomics Use Case</a></li>
<li class="tocline2"><a href="#s8_3">8.3 Radiotherapy Research Use Case</a></li>
<li class="tocline2"><a href="#s8_4">8.4 Computational Network Biology</a></li>
<li class="tocline2"><a href="#s8_5">8.5 Safety Information Evaluation and Visual Exploration ("SIEVE")</a></li>
<li class="tocline2"><a href="#s8_6">8.6 Sampling Large RDF Graphs</a></li>
<li class="tocline2"><a href="#s8_7">8.7 Query Formulation Using Data Metrics</a></li>
<li class="tocline2"><a href="#s8_8">8.8 Data Providers</a></li>
<li class="tocline2"><a href="#s8_9">8.9 Data Catalogs</a></li>
<li class="tocline1"><a href="#references">9 References</a> </li>
</ul>
</div>
<!--
<h3 id="appendices">Appendices</h3>
<ul class="toc">
<li class="tocline2"><a href="#rdfbundles">A RDF Sources</a></li>
<li class="tocline2"><a href="#chlog">Change Log</a></li>
</ul>
-->
<hr />
<h2 id="introduction">1 Introduction</h2>
<p>Big Data presents an exciting opportunity to pursue large-scale analyses over collections of data in order to uncover valuable insights across a myriad of fields and disciplines. Yet, as more and more data is made available, researchers are finding it increasingly difficult to discover and reuse these data. One problem is that <strong>data are insufficiently described</strong> to understand what they are or how they were produced. A second issue is that <strong>no single vocabulary provides all key metadata fields</strong> required to support basic scientific use cases. A third issue is that <strong>data catalogs and data repositories all use different metadata standards</strong>, if they use any standard at all, and this prevents easy search and aggregation of data. Therefore, we need a guide to indicate what are the essential metadata, and the manner in which we can express it.</p>
<p>For the purposes of this document, we <strong>define a dataset</strong> as “A collection of data, available for access or download in one or more formats” [<a href="#DCAT">DCAT</a>]. For instance, a dataset may be generated as part of some scientific investigation, whether tabulated from observations, generated by an instrument, obtained via analysis, created through a mashup, or enhanced or changed in some manner. Research data are available in research publications and supplemental documents, in literature curated databases such as PharmGKB or the CTD, from research repositories such as BioMedCentral-BGI GigaScience [<a href="#GigaScience">GigaScience</a>], Nature Publishing Group’s Scientific Data [<a href="#ScientificData">ScientificData</a>], Dryad Digital Repository [<a href="#Dryad">Dryad</a>], FigShare [<a href="#FigShare">FigShare</a>], Harvard Dataverse [<a href="#Dataverse">Dataverse</a>]. Cross-repository access is possible through data catalogs such as Neuroscience Information Framework (NIF) [<a href="#NIF">NIF</a>], BioSharing [<a href="#BioSharing">BioSharing</a>], Identifiers.org Registry [<a href="#Identifiers.org">Identifiers.org</a>], Integbio Database / Catalog [<a href="#Integbio">Integbio</a>], Force11 [<a href="#Force11">Force11</a>], and CKAN’s datahub [<a href="#Datahub">Datahub</a>].</p>
<p>While several vocabularies are relevant in describing datasets, none are sufficient to completely provide the breadth of requirements identified in Health Care and the Life Sciences. The Dublin Core Metadata Initiative (DCMI) [<a href="#DCMI">DCMI</a>] Metadata Terms offers a broad set of types and relations for capturing document metadata. The Data Catalog Vocabulary (DCAT) [<a href="#DCAT">DCAT</a>] is used to describe datasets in catalogs, but does not deal with the issue of dataset evolution and versioning. <span class="todo">[add other metadata]</span> The Provenance Ontology (PROV) [<a href="#PROV">PROV</a>] can be used to capture information about entities, activities, and people involved in producing a piece of data or thing. The Vocabulary of Interlinked Datasets (VOID) [<a href="#VOID">VOID</a>] is an RDF Schema vocabulary for expressing metadata about RDF datasets. Schema.org has a limited proposal for dataset descriptions [<a href="#SCHEMA">SCHEMA</a>]. Thus, there is need to combine these vocabularies in a comprehensive manner that meets the needs of data registries, data producers, and data consumers.</p>
<p>Here we describe the efforts of a multi-stakeholder effort under the auspices of the W3C Semantic Web for Health Care and Life Sciences [<a href="#HCLS">HCLS</a>] Interest Group to produce a specification for the description of datasets that meets key functional requirements, uses existing vocabularies, and is expressed using the Resource Description Framework [<a href="#RDF">RDF</a>]. We discuss elements of data description including provenance and versioning, and describe how these can be used for data discovery, exchange, and query (with SPARQL). This then enables the retrieval and reuse of data to encourage reproducible science.</p>
<p>
Specifically, we provide
<ul>
<li>A community specification for describing datasets (Section 5) in RDF.</li>
<li>Detailed requirements (Section 4) that have been drawn from a wide range of use cases (Section 8).</li>
</ul>
</p>
<h2 id="scope">2 Scope</h2>
<p>This document focuses on common data elements and their value sets for the description of data. Although use cases are drawn from Health Care and the Life Sciences, this document will focus on requirements that are broadly applicable.</p>
<h2 id="conventions">3 Conventions</h2>
<p>The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.</p>
<p>RDF data in this document is expressed in Turtle [<a href="#ref-TURTLE">TURTLE</a>]. Queries on this data are expressed in SPARQL [<a href="#ref-SPARQL">SPARQL</a>]. The following namespace prefix bindings are assumed unless otherwise stated:</p>
<div style="text-align: center;">
<table style="border-collapse: collapse; border-color: #000000" border="1" cellpadding="5">
<tr><th>Prefix</th> <th>URI</th> <th>Description</th></tr>
<tr><td><code>cito:</code></td> <td><code>http://purl.org/spar/cito/</code></td> <td>Citation Typing Ontology</td></tr>
<tr><td><code>dcat:</code></td> <td><code>http://www.w3.org/ns/dcat#</code></td> <td>Data Catalog</td></tr>
<tr><td><code>dctypes:</code></td> <td><code>http://purl.org/dc/dcmitype/</code></td> <td>Dublin Core Metadata Types</td></tr>
<tr><td><code>dct:</code></td> <td><code>http://purl.org/dc/terms/</code></td> <td>Dublin Core Metadata Terms</td></tr>
<tr><td><code>foaf:</code></td> <td><code>http://xmlns.com/foaf/0.1/</code></td> <td>Friend-of-a-Friend</td></tr>
<tr><td><code>freq:</code></td> <td><code>http://purl.org/cld/freq/</code></td> <td>Collection Description Frequency Vocabulary</td></tr>
<tr><td><code>idot:</code></td> <td><code>http://identifiers.org/terms#</code></td> <td>Identifiers.org vocabulary</td></tr>
<tr><td><code>lexvo:</code></td> <td><code>http://lexvo.org/ontology#</code></td> <td>Lexical Vocabulary</td></tr>
<tr><td><code>pav:</code></td> <td><code>http://purl.org/pav/</code></td> <td>Provenance Authoring and Versioning ontology</td></tr>
<tr><td><code>prov:</code></td> <td><code>http://www.w3.org/ns/prov#</code></td> <td>PROV Ontology</td></tr>
<tr><td><code>rdf:</code></td> <td><code>http://www.w3.org/1999/02/22-rdf-syntax-ns#</code></td> <td>Resource Description Framework</td></tr>
<tr><td><code>rdfs:</code></td> <td><code>http://www.w3.org/2000/01/rdf-schema#</code></td> <td>RDF Schema</td></tr>
<tr><td><code>schemaorg:</code></td> <td><code>http://schema.org/</code></td></tr>
<tr><td><code>sd:</code></td> <td><code>http://www.w3.org/ns/sparql-service-description#</code></td> <td>SPARQL 1.1 Service Description</td></tr>
<tr><td><code>xsd:</code></td> <td><code>http://www.w3.org/2001/XMLSchema#</code></td> <td>XML Schema</td></tr>
<tr><td><code>vann:</code></td> <td><code>http://purl.org/vocab/vann/</code></td> <td>Vocabulary for Annotating vocabulary descriptions</td></tr>
</table>
</div>
<p>The letter sequence "###" is used to denote a placeholde for a positive integer (≥ 1) in examples. The integers that replace this surrogate are of type <a href="http://www.w3.org/TR/xmlschema11-2/#integer">xsd:integer</a>.</p>
<h2 id="functionalrequirements">4 Functional Requirement</h2>
<p>In this section we describe a set of essential functionality that the dataset descriptions must provide.</p>
<h3 id="s4_1">4.1 Dataset Identification, Description, Licensing, and Provenance</h3>
<p>High quality data descriptions are necessary to understand the nature and provenance of data including what the data is, the format the data is represented with, where the data can be retrieved from, what license is associated with the dataset, how it was generated, when it was generated, and who generated it. Importantly, such dataset descriptions should provide globally unique identifiers for specific versions and formats of datasets so that they may be used and referenced by others in downstream analyses.</p>
<h3 id="s4_2">4.2 Dataset Discovery (via Catalog)</h3>
<p>Data consumers need an easy mechanism to find datasets of interest. Data catalogs, identifier indices, and data standard registries such as BioSharing [<a href="#BioSharing">BioSharing</a>], identifiers.org Registry [<a href="#Identifiers.org">Identifiers.org</a>], datahub.io [<a href="#Datahub">Datahub</a>], NIF [<a href="#NIF">NIF</a>] are all important infrastructure that make it easier for users to find relevant data and even discuss their quality or utility. The availability of rich metadata will enable users to perform faceted searches of data items by placing restrictions the values of specific metadata fields.</p>
<h3 id="s4_3">4.3 Exchange of Dataset Descriptions</h3>
<p>Dataset descriptions should be in a standardized format that enables facile exchange between data providers. This allows data catalogs to synchronize and specialize their offerings. For example, BioDBCore [<a href="#BioDBCore">BioDBCore</a>, <a href="#Gaudetetal2010">Gaudet et al 2010</a>] is a community-driven effort overseen by the International Society for Biocuration [<a href="#ISB">ISB</a>], which defines a checklist, or minimum information standard, including the core attributes for the description of biological databases.</p>
<h3 id="s4_4">4.4 Dataset Linking</h3>
<p>The integration of data typically involves establishing a similarity between resources described in different datasets. Since datasets naturally evolve with time, it is important that information regarding the nature of the linking can be adequately captured e.g. the version and format of files and software used to generate the links.</p>
<p>A dataset may incorporate, or link to, data in other datasets, e.g. in the creation of a data mashup. Rather than repeating the dataset description of the source datasets, the derived dataset would link to the dataset description of the specific instance that they loaded.</p>
<h3 id="s4_5">4.5 Content Summary</h3>
<p>A breakdown of the entities and their relationships in a dataset is useful for communicating the structure and content of the dataset. This information can be used to enable dataset navigation, facilitate query construction and compare different versions of a dataset.</p>
<h3 id="s4_6">4.6 Monitoring of Dataset Changes</h3>
<p>The reproducibility of scientific investigations is often tied to the availability of the original data. However, as original dataset grow or change with time, it becomes important to understand what changes have occurred and how these may affect dependent analyses. A dataset description should provide the means by which to compare different dataset versions.</p>
<h2 id="datasetdescriptionlevels">5 Dataset Description Levels</h2>
<p>In this section, we describe the W3C HCLS recommendation for rich descriptions of datasets. There are three levels each covering a different type of resource describing the data:</p>
<ul>
<li><strong>Summary Level</strong>: The summary level provides a description of a dataset that is independent of a specific version or format.</li>
<li><strong>Version Level</strong>: The version level captures version-specific characteristics of a dataset.</li>
<li><strong>Distribution Level</strong>: The distribution level captures metadata about a specific form and version of a dataset.</li>
</ul>
<p style="text-align: center">
<img src="Figure1.png" alt="Figure 1: An overview of the relationships between dataset description levels. A single summary level description for a dataset will be related to one or more version level descriptions using dct:isVersionOf. Incremental versions may be specified using pav:previousVersion. Each version level description will be linked to one or more distribution level descriptions using dcat:distribution."></img><br />
Figure 1: An overview of the relationships between dataset description levels. A single summary level description for a dataset will be related to one or more version level descriptions using dct:isVersionOf. Incremental versions may be specified using pav:previousVersion. Each version level description will be linked to one or more distribution level descriptions using dcat:distribution.
</p>
<p>Note that a distribution is the realisation of the data in a specific file format. The different distributions of a dataset may not be semantically equivalent due to differences between the data formats. For example, a chemical dataset may be released as a rich RDF dataset but also as an <a href="http://en.wikipedia.org/wiki/Chemical_table_file#SDF">SD file</a>. In this case, the SD file will not contain the same data content as the RDF dataset.</p>
<p>For example, consider the ChEMBL dataset. It is manifested in 17 different versions, of which each version can be accessed in a variety of data formats. Thus, to be fully conformant with this specification, there would be one Summary Level description that is linked to 17 Version Level descriptions, and each of the Version Level descriptions would be linked to their corresponding Distribution Level descriptions. This is captured by the following example expressed in RDF/Turtle format:</p>
<pre>
# Summary Level Description
:chembl a dctypes:Dataset .
# Version Level Description
:chembl17 a dctypes:Dataset ;
dct:isVersionOf :chembl;
dcat:distribution :chembl17rdf,
:chembl17relational, :chembl17csv .
# Distribution Level Descriptions
:chembl17rdf a dcat:Distribution, void:Dataset .
:chembl17relational a dcat:Distribution .
:chembl17csv a dcat:Distribution .
</pre>
<p>The properties expected for each metadata profile are given in the table below.</p>
<p style="text-align: center">
<table>
<tr>
<td>Profile</td>
<td>Property</td>
<td>Value</td>
<td>Summary Level</td>
<td>Version Level</td>
<td>Distribution Level</td>
</tr>
<tr>
<td>Core Metadata</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>rdf:type</td>
<td>dct:Dataset | void:Dataset</td>
<td>MUST</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dct:title</td>
<td>rdfs:langString</td>
<td>MUST</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dct:identifier</td>
<td>xsd:string</td>
<td>MAY</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:alternative</td>
<td>rdfs:langString</td>
<td>MAY</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:description</td>
<td>rdfs:langString</td>
<td>MUST</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dct:created</td>
<td>xsd:dateTime</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>pav:createdOn</td>
<td>xsd:dateTime</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>pav:authoredOn</td>
<td>xsd:dateTime</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>pav:curatedOn</td>
<td>xsd:dateTime</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:creator</td>
<td>IRI or xsd:string</td>
<td>NEVER</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dct:contributor</td>
<td>IRI or xsd:string</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>pav:createdBy</td>
<td>IRI or xsd:string</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>pav:authoredBy</td>
<td>IRI or xsd:string</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>pav:curatedBy</td>
<td>IRI or xsd:string</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:issued</td>
<td>xsd:dateTime</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>dct:publisher</td>
<td>IRI</td>
<td>MUST</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>foaf:page</td>
<td>IRI</td>
<td>SHOULD</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>schemaorg:logo</td>
<td>IRI</td>
<td>SHOULD</td>
<td>SHOULD</td>
<td>SHOULD</td>
<td></td>
</tr>
<tr>
<td></td>
<td>dct:license</td>
<td>IRI</td>
<td>MUST</td>
<td>MUST</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dct:rights</td>
<td>rdfs:langString</td>
<td>MAY</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:language</td>
<td><a>http://lexvo.org/id/iso639-3/{tag}</a></td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>dcat:theme</td>
<td>skos:Concept</td>
<td>MAY</td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>dcat:keyword</td>
<td>xsd:string</td>
<td>MAY</td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>void:vocabulary</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>dct:conformsTo</td>
<td>IRI</td>
<td>NEVER</td>
<td>MAY</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>cito:citesAsAuthority</td>
<td>IRI</td>
<td>MAY</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:hasPart</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>void:subset</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td>Identifiers</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>idot:preferredPrefix</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>idot:alternatePrefix</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>idot:identifierPattern</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>void:uriRegexPattern</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>idot:accessPattern</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>idot:exampleIdentifier</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:exampleResource</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td>Provenance and Change</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>void:inDataset</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>pav:version</td>
<td>xsd:string</td>
<td>NEVER</td>
<td>MUST</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>dct:isVersionOf</td>
<td>IRI</td>
<td>NEVER</td>
<td>MUST</td>
<td>NEVER</td>
</tr>
<tr>
<td></td>
<td>pav:previousVersion</td>
<td>IRI</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>pav:hasCurrentVersion</td>
<td>IRI</td>
<td>MAY</td>
<td>NEVER</td>
<td>NEVER</td>
</tr>
<tr>
<td></td>
<td>dct:source</td>
<td>IRI</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>pav:retrievedFrom</td>
<td>IRI</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>prov:wasDerivedFrom</td>
<td>IRI</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>pav:createdWith</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dct:accrualPeriodicity</td>
<td>dctype:Frequency</td>
<td>SHOULD</td>
<td>NEVER</td>
<td>NEVER</td>
</tr>
<tr>
<td></td>
<td><span class="todo">???</span>:latencyOfChange</td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td>Availability/Distributions</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>dct:format</td>
<td>[ iana | edam | biosharing | xsd:String ]</td>
<td>NEVER</td>
<td>NEVER</td>
<td>MUST</td>
</tr>
<tr>
<td></td>
<td>dcat:distribution</td>
<td>[ a dcat:Distribution; dcat:downloadURL <uri>]</td>
<td>NEVER</td>
<td>SHOULD</td>
<td>NEVER</td>
</tr>
<tr>
<td></td>
<td>dcat:accessURL</td>
<td>IRI</td>
<td>MAY</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td></td>
<td>dcat:downloadURL</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:dataDump</td>
<td>IRI</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>dcat:landingPage</td>
<td>IRI</td>
<td>NEVER</td>
<td>MAY</td>
<td>MAY</td>
</tr>
<tr>
<td>Statistics</td>
<td></td>
<td></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr>
<td></td>
<td>void:triples</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:entities</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:distinctSubjects</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:properties</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:distinctObjects</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>###:distinctLiterals</td>
<td>xsd:integer</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>sd:namedGraph</td>
<td>sd:NamedGraph</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:propertyPartition</td>
<td>[void:property <IRI>; void:triples "###"^^xsd:integer]</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
<tr>
<td></td>
<td>void:classPartition</td>
<td>[void:class <IRI>; void:entities "###"^^xsd:integer]</td>
<td>NEVER</td>
<td>NEVER</td>
<td>SHOULD</td>
</tr>
</table>
</p>
<h2 id="metadataguidancenotes">6 Metadata Guidance Notes</h2>
<h3 id="s6_1">6.1 Literals</h3>
<h4>6.1.1 Dates</h4>
<p>Dates should be given as accurately as possible. It is recommended that <a href="http://www.w3.org/TR/xmlschema-2/#dateTime">xsd:dateTime</a> is used (YYYY-MM-DDThh:mm:ss[Z], where [Z] is optionally specified as Z for UTC time or provided as a time zone offset e.g. -04:00 for EDT). In the case that the time is not precisely known, we recommend using the first valid value for the fields that are not known, i.e. the value ‘01’ for days and months, the value ‘00’ for hours, minutes and seconds.</p>
<h4>6.1.2 Values as Strings</h4>
<p>Values should be stated with a language tag, unless capturing an identifier or some structured value e.g. version identifier. Values may be captured in multiple languages.</p>
<h4>6.1.3 Values as Strings or IRIs</h4>
<p>For several predicates there is the choice of providing the value as an IRI or a string. We recommend that IRIs are used wherever possible.</p>
<h3 id="s6_2">6.2 Core Metadata</h3>
<h4>6.2.1 Dataset Identification and Declaration of Type</h4>
<p>It is a good general practice that data represented using Semantic Web technologies be typed as an instance of some class of entities. All datasets are typed as dctypes:Dataset with the exception of RDF formatted datasets which are typed as an instance of a void:Dataset.</p>
<pre>
:chembl
rdf:type dctypes:Dataset .
</pre>
<h4>6.2.2 Title, Alternative Names, and Short Code</h4>
<p>A dataset can be known by several names. At least one of these must be provided for use in human interfaces for applications using the dataset. It is expected that the preferred – commonly used – name for a dataset will be provided using the dct:title term, while alternative forms of the name, or older names for the dataset, may be given using the dct:alternative property. We also recommend that data providers specify exactly one short-code for a dataset using the dct:identifier property and that it should be in lower case.</p>
<p>For example, to provide a title, alternative name, and identier for the ChEMBL dataset:</p>
<pre>
:chembl
dct:title “ChEMBL”@en ;
dct:alternative “ChEMBLdb”@en ;
dct:identifier “chembl”@en .
</pre>
<h4>6.2.3 Description</h4>
<p>It is expected that the description will be the same as the text that appears on the web page for the dataset (see below). It should be a few paragraphs that explain to a domain expert the contents of the dataset. Where available, the description should be provided in alternative languages, each with the appropriate language tag. For instance, to specify a description for the chembl dataset:</p>
<pre>
:chembl
dct:description “ChEMBL is a database of bioactive compounds, their quantitative properties and bioactivities (binding constants, pharmacology and ADMET, etc). The data is abstracted and curated from the primary scientific literature.”@en .
</pre>
<h4>6.2.3 Dates of Creation and Issuance</h4>
<p>It is essential to know when a dataset came into existence. However for some datasets all that is known is when it was made public. Therefore, it is recommended that at least one of the creation or issued dates are provided. Both MAY be provided. For versioned or distribution dataset descriptions, state the date the dataset was generated using dct:created and/or the date the dataset was made public using dct:created. Both properties are restricted to a <a href="http://www.w3.org/TR/xmlschema-2/#dateTime">xsd:dateTime</a> value. For instance, to specify that the chembl17 dataset (see section 6.4 for more details about :chembl17 declaration) was issued on 29 August 2013:</p>
<pre>
:chembl17
dct:issued “2013-08-29T00:00:00”^^xsd:dateTime .
</pre>
<p>Note that since the time that the dataset was published is unknown, these are set to the beginning of the time period. If only the month and year had been known, the value would have been given as “2013-08-01T00:00:00”. Where the time is known, the timezone should be provided.</p>
<p>Other, more specific properties, e.g. the date the dataset was authored or curated, MAY be given using terms from the PAV ontology [<a href="#PAV">PAV</a>].</p>
<h4>6.2.4 Authorship, Creation, Curation</h4>
<p>Details of the individual or organisation responsible for creating a dataset MUST be provided using the dct:creator property. The value should be an IRI for the individual or organisation that can be resolved for more details. We recommend the use of ORCID ID [<a href="#ORCID">ORCID</a>] for researchers. The value may be a string if a suitable IRI cannot be supplied.</p>
<pre>
:chembl17
dct:creator <https://www.ebi.ac.uk/chembl/> .
:chembl_target_targetcmpt_linkset
pav:authoredBy <http://orcid.org/0000-0002-8011-0300> ;
</pre>
<p>Fine-grained attribution of creation events such as authoring or curation MAY additionally be supplied using the terms from the PAV ontology.</p>
<ul>
<li><strong>Authorship</strong>: State the author(s) with pav:authoredBy and linking to URIs for the authors. The date of authorship should be given using pav:authoredOn with a <a href="http://www.w3.org/TR/xmlschema-2/#dateTime">xsd:dateTime</a>.</li>
<li><strong>Creation</strong>: State the creator(s) with pav:createdBy and linking to URIs for the authors. The date of authorship should be given using pav:createdOn with a <a href="http://www.w3.org/TR/xmlschema-2/#dateTime">xsd:dateTime</a>.</li>
<li><strong>Curation</strong>: State the author(s) with pav:curatedBy and linking to URIs for the authors. The date of authorship should be given using pav:curatedOn with a <a href="http://www.w3.org/TR/xmlschema-2/#dateTime">xsd:dateTime</a>.</li>
</ul>
<pre>
:chembl_target_targetcmpt_linkset
pav:authoredBy <http://orcid.org/0000-0002-8011-0300> ;
</pre>
<p>Declare the tool used to create the dataset using pav:createdWith linking to a URI representing the specific version of the tool.</p>
<h4>6.2.5 Publisher</h4>
<p>A link to the organisation responsible for publishing the dataset.</p>
<pre>
:chembl
dct:publisher <http://www.ebi.ac.uk> .
</pre>
<h4>6.2.6 Webpage and Logo</h4>
<p>A link to the human-oriented web page for the dataset is provided using the foaf:page property [<a href="#FOAF">FOAF</a>]. Due to the inverse-functional properties, foaf:homepage must not be used as it could result in the data about different versions of a dataset being combined. For instance, to specify that the chembl dataset is accessible at [<a href="#ChEMBL">ChEMBL</a>]:</p>
<p>A link to an image file containing the logo for the dataset is provided using the schemaorg:logo property [<a href="#SCHEMAORG">SCHEMA.ORG</a>]. Due to the inverse-function properties of foaf:logo must not be used.</p>
<pre>
:chembl
foaf:page <http://www.ebi.ac.uk/chembl/> ;
schemaorg:logo <http://www.ebi.ac.uk/rdf/sites/ebi.ac.uk.rdf/files/resize/images/rdf/chembl_service_logo-146x48.gif> .
</pre>
<h4>6.2.7 Keywords</h4>
<p>Keywords and topics of coverage of the dataset MAY be given. It is recommended that such terms are drawn from a suitable domain vocabulary and declared using the dcat:theme predicate. The dcat:keyword predicate is provided for the rare occasions where there is not a suitable term in a vocabulary.</p>
<pre>
@prefix ncit: <http://ncicb.nci.nih.gov/xml/owl/EVS/Thesaurus.owl#>
:chembl
dcat:theme ncit:C48807. # chemical
dcat:keyword “chemical”^^xsd:string, “assay”^^xsd:string .
</pre>
<p>Note that the ncit namespace refers to the National Cancer Institute (NCI) Thesaurus.</p>
<h4>6.2.8 Licensing and Rights</h4>
<p>The license under which the data is published MUST be stated to enable the reuse of the data. The details of the license may be summarised in a rights statement.</p>
<pre>
:chembl
dct:license <http://creativecommons.org/licenses/by-sa/3.0/> .
</pre>
<h4>6.2.9 Language</h4>
<p>Declare the languages the data is published in using dct:language and values taken from the Lexvo.org Ontology [<a href="#Lexvo">Lexvo</a>].</p>
<pre>
:chembl
dct:language <http://lexvo.org/id/iso639-3/en> .
</pre>
<h4>6.2.10 Literature Reference</h4>
<p>Provide a literature reference as a CiTO entry [<a href="#CiTO">CiTO</a>] using cito:citesAsAuthority property. For example, for the :chembl dataset the literature reference is [<a href="#Gaultonetal2011">Gaulton et al 2011</a>].</p>
<pre>
:chembl
cito:citesAsAuthority <http://nar.oxfordjournals.org/content/40/D1/D1100>
</pre>
<h4>6.2.11 Vocabularies</h4>
<p>RDF datasets use one or more RDFS vocabularies or OWL ontologies to represent the data. It is recommended that the distribution level metadata, for RDF datasets, should state what terminologies were used in the dataset using void:vocabulary [<a href="#VOID">VOID</a>]. The value has to be a URI identifying the terminology used.</p>
<pre>
:chembl
void:vocabulary <> .
</pre>
<h4>6.2.12 Conformance</h4>
<p>Indicate that the dataset conforms to a particular format or standard using the dct:conformsTo property. For example:</p>
<pre>
:dataset
dct:conformsTo <span class="todo"><></span> .
</pre>
<h4>6.2.13 Subsets</h4>