- May 2015
-
www.semantic-web-journal.net www.semantic-web-journal.net
-
The model uses the main clas- ses of Darwin Core.
Figure 1 doesn't make it evident which classes are from DwC and which ones are not. This should be made visually clear in the figure, and mentioned in the figure legend.
-
with an outline of the model suggest- ed by Richard Pyle
Acknowledgement of individual contributors and groups of contributors (such as mailing lists) whose work didn't rise to the level of co-authorship should all be moved to the Acknowledgements section. In the middle of the text they distract.
-
This approach differs significantly from that taken in the
How so? It it's important, it should be expanded.
-
DSW sees class instances as nodes that group related properties rather than as enti ties that are heavily constrained ontologically
I'm not sure I understand what is meant by this. I question though whether the authors have as much choice in "seeing" class instances as the sentence could be made out to say. If this is an OWL ontology (which I understand it is), and instance data are expressed in OWL, then class instances are owl:Individuals, which are entities, not "nodes". I.e., this statement makes me wonder whether the authors' view of their instance data might be in conflict with the very entailments of their OWL ontology.
-
is to facilitate the linking of real data
In contrast to unreal or theoretical data? Probably not. Express more directly what the motivation or driving goal is.
-
In many cases, URIs identify real resources in the wild, although example triples composed of those URIs are not necessarily asserted there.
This doesn't make much sense to me. What are the authors trying to say, and why is it important?
-
Those instances can serve as anchor points to which resources outside the biodi- versity community may be linked.
Need to explain why this can be a useful thing to do.
-
The Darwin - SW model
The readership of SWJ should not be assumed to have any biodiversity domain knowledge, and hence I suggest that as the first subsection the authors briefly introduce the salient aspects of a canonical biodiversity observation.
-
Instances of key classes within the biodiversity domain (such as organisms, specimens, and taxa) within or among institutions can be linked using DSW - defined object properties that have no ana- logues in D arwin Core.
This will likely be very confusing to anyone not intimately familiar with biodiversity domain knowledge. Also, it's not obvious why these properties are missing from DwC.
-
the choice was made
Assuming that it was the authors who made this choice, use active voice.
-
However, even with the RDF guide, Darwin Core does not provide object properties to link instances of its main classes
This needs expanding - obviously it's one of the main motivating use cases for DSW.
-
It was developed in 2010 - 11 to meet an immediate need in the Bioimages 22 and Xmalesia 23 projects
I suggest to expand this - it could for the motivating use-cases from which to hang and unroll the unmet challenges, and consequently the goals behind developing DSW.
-
Introduction
This section needs to be brought into a proper narrative. In its current form, especially following there is much to draw from the text is more like an enumerated sequence (which isn't even strictly chronological) and lacks a narrative thread that keeps a reader oriented.
-
seria l- ized as
expressed as?
-
and lack of funding made their maintenance difficult .
Doesn't this apply to any standard, or more broadly informatics product, including the presented here?
-
was known
became known?
-
A central part of this architecture was the creation of an ontol- ogy to facilitate the integration of standards .
What standards? Surely not any standard?
-
However, progress towards exposing biodiversity data as RDF has been ham- pered by the lack of a consen sus graph - based model for the biodiversity domain
Isn't the reason more the debate over consensus semantics?
Tags
Annotators
URL
-