/[volute]/trunk/projects/semantics/veps/VEP-003.txt
ViewVC logotype

Contents of /trunk/projects/semantics/veps/VEP-003.txt

Parent Directory Parent Directory | Revision Log Revision Log


Revision 5843 - (show annotations)
Mon Jul 13 08:31:38 2020 UTC (11 months, 1 week ago) by msdemlei
File MIME type: text/plain
File size: 3278 byte(s)
misc VEPs: Fixed vocabulary URIs, put a suggestive default into the template.

1 Vocabulary: http://www.ivoa.net/rdf/datalink/core
2 Author: Fran├žois Bonnarel, Markus Demleitner, msdemlei@ari.uni-heidelberg.de
3 Date: 2019-12-06
4 Superceded-by: VEP-004
5
6 New Term: sibling
7 Action: Addition
8 Label: Sibling Data
9 Description: Data products derived from the same progenitor as #this.
10 This could be a lightcure for an object catalog derived from repeated
11 observations, the dataset processed using a different pipeline, or the
12 like.
13 Used-in:
14 http://dc.g-vo.org/gaia/q2/tsdl/dlmeta?ID=ivo://org.gavo.dc/~?gaia/q2/199286482883072/BP
15 This is GAVO's rendition of the Gaia DR2 epoch photometry, where
16 users retrieve a time series in a specific band; the time series
17 in the other bands are the siblings of that.
18
19 Rationale:
20 It is fairly common in complex pipelines that multiple data products
21 result from a single observation. Often, this is true even in a
22 single pipeline step, and hence the data products are not in a
23 progenitor-derivation relationship. Still, researchers will want to
24 know about these data products; for instance, while exploring a source
25 in Gaia, a quick way to access epoch photometry or the RP/BP spectra
26 is obviously valuable; such artefacts are not really progenitors of
27 the catalog entry, though. In such cases, #sibling (or perhaps one of
28 its future child terms) should be used.
29
30 Clients should offer #sibling links in a context of scientific
31 exploitation of the dataset (as opposed to, say, debugging).
32
33 Discussion:
34 In the discussion, it was the need for the concept as such ("other
35 things that were produced from the observations that led up to #this")
36 was not disputed, though the discussion was somewhat delayed by
37 an investigation of possible shortcomings in the datalink data model
38 (http://mail.ivoa.net/pipermail/dal/2019-December/008248.html) and
39 whether additional cases should or should not be included in it
40 (http://mail.ivoa.net/pipermail/dal/2020-February/008262.html).
41
42 However, the main points of contention were the choice of the term and
43 label ("sibling"). Objections included that astronomers might not
44 understand the provenance-inspired nomenclature, that a very rough
45 view of provenance must be adopted to actually talk about siblings
46 (because, really, #this and the #sibling items just share common
47 ancestors, not (necessarily) the parents), or that it is confusing to
48 define, say, a spectrum to be the sibling of a catalogue row
49 (http://mail.ivoa.net/pipermail/semantics/2020-May/002700.html).
50
51 Possible alternatives investigated include #see-also (which was
52 rejected as being too general), #co-generated (which was disliked
53 because the implication that the two artefacts were built at the same
54 time by the same processing step is even stronger than with #sibling),
55 and #coderived (which wide acceptance but was strongly rejected by one
56 party arguing it would strongly distort the meaning of "derived".
57
58 In the end, #sibling was accepted as being acceptable after
59 a splinter discussion during the May 2020 Virtual Interop, but the
60 term #cogenerated was considered preferable, mainly because it
61 better showed that the concept applies to elements in long
62 provenance trees even if they are only related through distant
63 ancestors.
64
65

msdemlei@ari.uni-heidelberg.de
ViewVC Help
Powered by ViewVC 1.1.26