W3C home > Mailing lists > Public > public-gld-wg@w3.org > February 2012

ISSUE-33 (Slices): Collections of observations and well-formedness of slices [Data Cube Vocabulary]

From: Government Linked Data Working Group Issue Tracker <sysbot+tracker@w3.org>
Date: Fri, 17 Feb 2012 16:38:00 +0000
Message-Id: <E1RyQoq-0000Qu-Uc@tibor.w3.org>
To: public-gld-wg@w3.org
ISSUE-33 (Slices): Collections of observations and well-formedness of slices [Data Cube Vocabulary]

http://www.w3.org/2011/gld/track/issues/33

Raised by: Dave Reynolds
On product: Data Cube Vocabulary

The qb:Slice structure is used to collect all qb:Observations that have a given fixed set of values for a subset of the cube's dimensions. This is used to support abbreviated format and to indicate to data consumers how best to present the cube. For example, many statistical datasets are organized as time series so that the slices fix every dimension except time.

Thus implicitly for a given qb:SliceKey there is a disjoint cover of qb:Slices (every observation is in one and only one slice with that key).

In practice we have observed a number groups using, or suggesting use slices as a way to group more arbitrary subsets of the observations. Specific uses cases include collecting all the "latest" observations  in a continuously updated data set (bathing water quality use case), or collecting all observations for a particular subset of time or space for efficiency of access (air quality use case).

(1) Is it legitimate to use qb:Slice without a well defined qb:SliceKey for these more arbitrary collections of observations?

(2) If not should Data Cube include a qb:Collection mechanism to support such use cases?

(3) Should the domain constraint on qb:observation (currently rdfs:domain qb:Slice) be lifted to it to be used for such collections.
Received on Friday, 17 February 2012 16:38:05 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:32:35 UTC