DIMENSION in the tests

[ArthurChapman]

We have a field in the Tests called DIMENSION (NAME, SPACE, TIME, OTHER)

All the NAME tests use the Darwin Core Class: Taxon - except one (#106 ) that is Taxon and Identification,

All the TIME tests use the Darwin Core Class: Event

All the SPACE tests use the Darwin Core Class: Location

The OTHER - five use the Darwin Core Class: Occurrence
- eight use the Darwin Core Class: Record_level Terms

#72 a NOTIFICATION that has 3 Dimensions (Space, Time, Name) and refer to the Darwin Core Class "Record Level". This could better be OTHER

The MEASURE tests which have Darwin Core Class :"All" except for one (#140 ) which as Darwin Core Class: Event

I think this gives us a lot of Duplication - instead of DIMENSION (NAME, SPACE, TIME, OTHER) which have been great as a working tool - and we can still use those labels in GitHub, I see little value in including the field in the tests - and can just use the Darwin Core Classes (Taxon, Event, Location, Occurrence, Identification, Record_level Terms, and All)

[ArthurChapman]

This would also save us having to define DIMENSION, NAME, SPACE, TIME, OTHER in the Vocabulary

[chicoreus]

The name "Dimension" is confusing, as a fundamental concept in the Framework is Dimension, meaning data quality dimension, while the values listed here are categories of Information Elements. The labels applied to the tests in the github issue tracker with these values (NAME, TIME, etc) have definitely been valuable in clustering tests of interest as we've been discussing them, so perhaps there's an issue here to put against TG1 to provide a framework element to categorize Information Elements (i.e. an Information Element having properties of label (e.g. Coordinate), specific terms (e.g. dwc:decimalLatitude, dwc:decimalLongitude, dwc:geodeticDatum, dwc:coordinateUncertaintyInMeters, dwc:coordinatePrecision), and category (e.g. SPACE).

This would allow TG2 to define InformationElement category values such as SCINAME, TIME, SPACE, OTHERDWC).

[tucotuco]

Might there be merit in using Darwin Core classes among the clustering categories?

…

On Fri, Sep 7, 2018 at 2:56 AM Paul J. Morris ***@***.***> wrote: The name "Dimension" is confusing, as a fundamental concept in the Framework is Dimension, meaning data quality dimension, while the values listed here are categories of Information Elements. The labels applied to the tests in the github issue tracker with these values (NAME, TIME, etc) have definitely been valuable in clustering tests of interest as we've been discussing them, so perhaps there's an issue here to put against TG1 to provide a framework element to categorize Information Elements (i.e. an Information Element having properties of label (e.g. Coordinate), specific terms (e.g. dwc:decimalLatitude, dwc:decimalLongitude, dwc:geodeticDatum, dwc:coordinateUncertaintyInMeters, dwc:coordinatePrecision), and category (e.g. SPACE). This would allow TG2 to define InformationElement category values such as SCINAME, TIME, SPACE, OTHERDWC). — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#166 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAcP6_oF6JcVgmwGtt3kLuDvazCGWrvZks5uYTeUgaJpZM4WcOEo> .

[ArthurChapman]

@chicoreus I agree that if we keep them we should change DIMENSION to something else so as to not confuse with Framework. I also find TIME confusing as it refers to Event rather than time and in fact we have very few references to dwc:time.

@tucotuco for me - yes - I believe we can do without and just follow Darwin Core Classes. I like keeping at simple as possible. I would like to hear what @Tasilee and others think

[Tasilee]

As long as we can easily categorise each test as NAME, SPACE, TIME and OTHER from Darwin Core Class, then those terms are redundant.