Should observationType animal be broadened to species

[peterdesmet]

Currently, biological observations are indicated as animal, because camera traps are typically used for animal observations. I'm wondering however if it is a good idea to limit this from the start or if we should use a broader term like species to allow for plant (phenology) and fungi observation.

camtrap-dp/observations-table-schema.json

Lines 84 to 100 in 868c7eb

	{
	"name": "observationType",
	"description": "Type of the observation. All categories in this vocabulary have to be understandable from an AI point of view. `unknown` describes classifications with a `classificationProbability` below some predefined threshold i.e. neither humans nor AI can say what was recorded.",
	"type": "string",
	"constraints": {
	"required": true,
	"enum": [
	"animal",
	"human",
	"vehicle",
	"blank",
	"unknown",
	"unclassified"
	]
	},
	"example": "animal"
	},

Note that "All categories in this vocabulary have to be understandable from an AI point of view."

Alternatively, we keep animal but allow plant and fungus. Personally, I'm less in favour of this, since from a documentation standpoint it makes sense to refer to a single term rather than three:

Observations can mark non-animalspecies events (camera setup, human, blank) or one or more animalspecies observations (observationType = animalspecies)

Users can always filter further based on scientific name and provided taxonomy.

[peterdesmet]

Discussed with @kbubnicki. The categories in observationType are informed by AI classification. The current list represents the broad classification currently supported by AI. It is therefore useful to keep animal over a lumped species, since AI will likely continue to classify animals differently from humans, plants, etc.

We agreed that with new AI possibilities, the list can be extended. For example with plant and fungus. Since that change is currently not backed up by a use case, we have moved this to 1.1.