 |
| Titel |
CHARMe Commentary metadata for Climate Science: collecting, linking and sharing user feedback on climate datasets |
| VerfasserIn |
Jon Blower, Bryan Lawrence, Philip Kershaw, Maurizio Nagni |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250098856
|
| Publikation (Nr.) |
 EGU/EGU2014-14574.pdf |
|
|
|
|
|
| Zusammenfassung |
| The research process can be thought of as an iterative activity, initiated based on prior domain
knowledge, as well on a number of external inputs, and producing a range of outputs
including datasets, studies and peer reviewed publications. These outputs may describe the
problem under study, the methodology used, the results obtained, etc. In any new publication,
the author may cite or comment other papers or datasets in order to support their
research hypothesis. However, as their work progresses, the researcher may draw
from many other latent channels of information. These could include for example,
a private conversation following a lecture or during a social dinner; an opinion
expressed concerning some significant event such as an earthquake or for example
a satellite failure. In addition, other sources of information of grey literature are
important public such as informal papers such as the arxiv deposit, reports and
studies.
The climate science community is not an exception to this pattern; the CHARMe project,
funded under the European FP7 framework, is developing an online system for collecting and
sharing user feedback on climate datasets. This is to help users judge how suitable such
climate data are for an intended application. The user feedback could be comments about
assessments, citations, or provenance of the dataset, or other information such as descriptions
of uncertainty or data quality. We define this as a distinct category of metadata
called Commentary or C-metadata. We link C-metadata with target climate datasets
using a Linked Data approach via the Open Annotation data model. In the context
of Linked Data, C-metadata plays the role of a resource which, depending on its
nature, may be accessed as simple text or as more structured content. The project is
implementing a range of software tools to create, search or visualize C-metadata including a
JavaScript plugin enabling this functionality to be integrated in situ with data provider
portals.
Since commentary metadata may originate from a range of sources, moderation of this
information will become a crucial issue. If the project is successful, expert human moderation
(analogous to peer-review) will become impracticable as annotation numbers increase, and
some combination of algorithmic and crowd-sourced evaluation of commentary
metadata will be necessary. To that end, future work will need to extend work under
development to enable access control and checking of inputs, to deal with scale. |
|
|
|
|
|
|