Data usability


Computing trust on a given level of Data Usability is envisionned here as a cross-cutting activity, encompassing :
- the Data Modelling and its fit for purpose towards some given end-user applications
- the ability to estimate and record data quality values, especially through uncertainty calculation and modelling
- the ability to confidently exploit, update and keep track of these quality descriptions, notably within reliable metadata records
Computing trust in this context is oftenly related to :
- the requirement for a data accuracy management strategy
- the availability of uncertainty estimates, associated with the chain of measurements and processings of such data
- the ability to communicate and make use of such estimates in an interoperable manner


During the AIP-3 process, the Data Harmonization Working Group discussed some research work and best practices.
Out of the different participating organisations, some consolidated topics are tentatively presented here below.

Quality Assurance (e.g. QA4EO for GEOSS)

QA4EO was established by CEOS as a contribution to facilitate the GEO vision for a Global Earth Observation System of Systems (GEOSS), initially relying on the Space Domain knowledge coming from calibration and validation experts.
There is a good entry point to the topic of ‘QA4EO for GEO’ on the GEOSS Best Practices Wiki, Data Quality Assurance pages.

Data Usability (e.g. EC Framework Programmes and INSPIRE)

An approach (Article and concept diagram) discussed in the GIS4EU project (INSPIRE Data Harmonization context):

The NetMar project is addressing standards for data and metadata formats, as well as exchange protocols, as the first steps to bridge existing marine data systems. The next step is looking at the semantics of the services, including an uncertainty model, to allow transparent computer-based discovery.

The GIGAS project.
GIGAS results presneted to AIP-3 had a focus on ISO19156 Observations & Measurement
This ISO19156 standard (and ISO and OGC standards, more generally) have been discussed in regard of the need to pay attention to making standards sensibly modular, that several aspects in 19156 are very generic in nature, and have relevance across a large set of uses. Specifically, the question of modularity to 'import' a standard like UncertML was raised !
The ISO19156 rationale for including quality information could be made more strongly and more explicitly (like in Section B5.1 on discovery and use).

There is also a Future Sensor Web forum group within INSPIRE that examines uncertainty.

CSIRO and the Water for a Healthy Country Flagship

As part of the Water for a Healthy Country Flagship and the Intelligent Island Program, the CSIRO has formed a partnership with the Rensselaer Polytechnic Institute to develop a web-based provenance management system for the water domain. Initially, the focus will be on the development of use cases that are applicable to the flow forecast system that the CSIRO has developed (soon to be registered in GEOSS).
The use cases can be viewed here.

DHWG Engineering Report - Proposed Table of Content

(based on the outcomes and results gathered from AIP-3 participants, to structure the DHWG Engineering Report)


- Defining the scope of data, products and metadata harmonization within GEOSS
- Resuming the way that topic is also adressed in other contexts (INSPIRE initiative, CSIRO projects, Projects under the EC Framework Programmes, ...)

Data Harmonization requirements

Out of CSIRO, PML, Aston Univ., Munster Univ., GIGAS contributions...

Quality assurance: which encoding standards, for which purposes

Out of QA4EO, Aston Univ., GIGAS contributions

Other chapters...

Metadata: coordination with other AIP WGs
Ontologies & registers: coordination with other AIP WGs
Ideas ?


Identified requirements reviews:
- Through AIP-3, the uncertainty quantification was suggested as the only principled way of combining imperfect information from multiple sources.
- Defining the universe of discourse (what we are able to represent and describe in our model), when dealing with such a subject matter, is also raised as a requirement.
- ...

Recommendations for the Task Forces:
- GEOSS Common Infrastructure TF
- Data Sharing TF
- ...


Hervé Caumont,
Nov 10, 2010, 8:00 AM