(minimum) bounding box and extent concepts not in MLGT

[heidivanparys]

The concepts "(minimum) bounding box" and "extent" cannot be found in GeoLexica, but those concepts are an important part of the geographic information vocabulary.

Problem seen in https://github.com/opengeospatial/ogcapi-common/issues/261 and also seen in the project on the Danish profile of DCAT-AP.

[PeterParslow]

There are a number of definitions of 'bounding box' on the ISO, from computer graphics & manufacturing. They are nearly what we are looking for, but it is surprising that we haven't defined the term anywhere ourselves - perhaps we considered a dictionary word/phrase that didn't need special definition?

ISO 19107:2019 defines the interface class "Envelope" and states "Envelope is often referred to as a minimum bounding box or rectangle" & we use the phrase quite extensively.

ISO 19115-1:2014 defines classes for EX_GeographicBoundingBox (and EX_BoundingPolygon)

I think this is an example of the question: what is the distinction between phrases/words defined as terms and those defined as classes? Perhaps the ISO 19103 dictionary of useful classes is actually a (fledgling) catalogue of concepts? @ReesePlews , @jetgeo , @WG1, @heidivanparys

[heidivanparys]

There are a number of definitions of 'bounding box' on the ISO, from computer graphics & manufacturing. They are nearly what we are looking for, but it is surprising that we haven't defined the term anywhere ourselves - perhaps we considered a dictionary word/phrase that didn't need special definition?

I would expect this to come from mathematics (@busstoptaktik?):

• minimum bounding rectangle (note the interesting link in the references: Unlocking the Mysteries of the Bounding Box)
• minimal enclosing circle

Calculation of it seems to belong to a field called "computational geometry", see e.g. https://cs.au.dk/~gerth/cg10/project3.html and https://ojs.cvut.cz/ojs/index.php/gi/article/download/gi.3.2/2563

I think this is an example of the question: what is the distinction between phrases/words defined as terms and those defined as classes? Perhaps the ISO 19103 dictionary of useful classes is actually a (fledgling) catalogue of concepts?

We are often mixing up describing concepts (terms, definitions, notes, anything that explains what we mean when we use a certain term) and describing data structures for storing and exchanging data. See e.g. https://gitlab.ogc.org/ogc/conceptual-modeling-discussion-paper/-/issues/8 (OGC Architecture DWG restricted issue) and https://www.brsolutions.com/concept-model-vs-data-model/.

We need both, I think:

• a terminology
• classes that reuse that terminology, for reuse in platform-independent data models

Ideally, we would then have some kind of automated way to copy definitions etc. from the terminology resource to the classes (if the terminological entries have a unique identifier and are machine-readable, that is possible).

[heidivanparys]

This is being discussed in OGC in the Architecture DWG, in the conceptual modeling subgroup more specifically.

[ReesePlews]

thank you @heidivanparys these are examples of critical terminology concepts that need to be added to the tc211 vocabulary. they should be discussed more in the Terminology PWI once that is underway. i hope you and your colleagues will be able to join that work.

[ogcscotts]

The OGC Technical Committee is currently drafting a conceptual model for "bounding box" potentially yo be published as an OGC Abstract Specification Topic.

[heidivanparys]

See also https://github.com/opengeospatial/Bounding-Box-Abstract-Specification.