RE: [BP] EO-QB meets BP

All good. Thanks Jeremy.

From: Jeremy Tandy []
Sent: Thursday, 6 April 2017 10:00 PM
To: Kerry Taylor <>; Linda van den Brink <>; Payam Barnaghi <>; Bill Roberts <>
Cc: Zhiduo Zhang <>; Samuel Toyer <>; Dmitry Brizhinev <>;
Subject: Re: [BP] EO-QB meets BP

And continuing ...

Point 2.

- I've added both EO-QB and QB4ST to the local biblio so we can reference them
- Added the following (the underlined bit) to BP 6:

> Standard data encodings are available for time-series data, including: [TIMESERIESML] for [GML], plus [COVERAGE-JSON] and [SENSORTHINGS] for JSON. [VOCAB-DATA-CUBE] provides a generic mechanism to express well-structured data, such as timeseries, in RDF. [EO-QB] and [QB4ST] (developed alongside this best practice Note within the Spatial Data on the Web Working Group<>) illustrate how [VOCAB-DATA-CUBE] may be used in this way.

... not suggesting that they're best practice, just illustrative of using RDF Data Cube.

Point 3. [CC/ Bill again.]

Section 6 doesn't mention any implementation approaches - and I'd rather keep it that way. There are lots beyond Cov-JSON and EO-QB. Instead, I wonder if Bill can find some way to stitch a reference to coverage data models and encodings into his edits of BP10? Again, I've added a reminder to the detailed plan [1].

Point 4.



See PR 668 [2] for the changes. Now merged.




On Thu, 6 Apr 2017 at 11:47 Jeremy Tandy <<>> wrote:
Bill - flagging point 1 of Kerry's email. This is something you might consider including in the edits on BP10. I've added a reminder to the detailed plan [1].

Kerry. I'm working through the other points and will update shortly.



On Tue, 4 Apr 2017 at 14:00 Kerry Taylor <<>> wrote:
Dear  BP editors,
This relates to eo-qb issue-107 and issue-100.

I have revised EO-QB to make reference to its application of best practices, you might like to have a look.

As I did it, I spotted  a few places where BP might like to reference EO-QB as  examples.

1.  Balancing quality and cost
BP Encoding spatial data needs an example for "Balancing quality and cost.” “For example, if using a Linked Data approach, one option is to keep all data in a triple store; but hybrid approaches are also possible, “ … This EO-QB could be the missing example,  Say…

In EO-QB  satellite imagery metadata is stored in a triple store but observational data is stored in HDF5  and served through DGGS and query processing middleware that materializes RDF triples only in response to a SPARQL query for them.

2.BP change over time
“Best Practice 6: Describe properties that change over time” I suppose EO-QB is an example of approach 3 “capturing a time-series of data values within an attribute of the spatial thing.” Although in eoqb he spatial thing itself (earth surface) is never really mentioned…  I suggest “EO-QB” should be added to the list of includings in the following (although it is not quite a standard itself, but it uses the RDF datacube standard to do it)  “Standard data encodings are available for time-series data, including: [TIMESERIESML<>] for [GML<>], plus [COVERAGE-JSON<>] and [SENSORTHINGS<>] for JSON.”

3. section 6
“Coverages: describing properties that vary with location (and time)" should reference EO-QB and COV-JSON.
Note that COV_JSON uses the domain/range/range metadata  model explicitly whereas in EO-QB this model is  obscured by the primary  conceptualisation as  a  3-dimensional Data Cube.

4. crossreferencing to UCR

BP doc should crossreference req  from


Received on Friday, 7 April 2017 03:54:34 UTC