New class: Sampling Process,

[dr-shorthair]

The OBI/BCO classes observing process and specimen collection process are immediate sub-classes of OBI's planned process.

However, specimen collection is just one kind of sampling activity. Other kinds are

• specimen processing (dissecting, splitting, grinding, dissolving, sieving)
• getting a sample of a population
• selecting or marking out fiat sites (transects, stations, etc)

So there could be an intermediate class in the subsumption hierarchy, between planned-process and specimen-collection-process.

Note that the primary intention of any of these sampling processes (activities) is to generate one or more new or transformed samples, where sample is understood more broadly than 'specimen' which is just a material sample.

This can be contrasted with observing process where the primary intention and result is data - i.e. tokens representing numerical, categorical, presence/absence values.

[ramonawalls]

STATO already has a class for "statistical sampling process" which I think can be reused for generalized sampling (see http://purl.obolibrary.org/obo/STATO_0000502).

OBI has a class for material sampling process (http://purl.obolibrary.org/obo/OBI_0000744) and material sample (http://purl.obolibrary.org/obo/OBI_0000747) which are subclasses of specimen collection process and specimen, respectively, but they were never completed.

I am going to create an issue in OBI, to see if these terms should be finalized there and imported into BCO, because I expect they are needed outside biodiversity studies.

[pbuttigieg]

We should consider a pattern for how STATO relates to physical processes. My feeling is that it should define the plans that govern sampling planned processes. Thus one or more instances of a material sampling process can fulfil a statistical sampling plan. This would make things more modular and prevent the autophagic import issues we face from time to time.

Also, I'd be reluctant to classify the marking out of sites as sampling.

[ramonawalls]

See issue obi-ontology/obi#969

[dr-shorthair]

I'd be reluctant to classify the marking out of sites as sampling.

Why? This seems analogous to the statistical sampling activity.

Marking out a site is an activity to select and delineate a subset of an ecosystem(s) or zone(s) with the intention that it is representative of the larger thing, in some way specified in the design.

[pbuttigieg]

If the object being sampled is the larger ecosystem that the sites are within (thus the sites themselves are sampled units) this works. Does that align with your understanding?

On Sat, 29 Sep 2018, 07:39 Simon Cox, notifications@github.com wrote:

I'd be reluctant to classify the marking out of sites as sampling.

Why? This seems analogous to the statistical sampling activity.

Marking out a site is an activity to select and delineate a subset of an ecosystem(s) or zone(s) with the intention that it is representative of the larger thing, in some way specified in the design.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/BiodiversityOntologies/bco/issues/87#issuecomment-425617947, or mute the thread https://github.com/notifications/unsubscribe-auth/ACK7Mq5LzdXVFRkXpN-qJnxW3r-LFzQ7ks5ufwemgaJpZM4Wxb-z .

[dr-shorthair]

Yes - that is exactly the case that I am thinking of.