Table 2. Evaluation criteria used to assign the relevance category to the datasets.

Classifier group	Classifier	Relevance assignment	Example
Main classifiers	Data type	High: Abundance, density, EBV genetic analysis.	L
		Moderate: Distribution, presence-absence.
		Low: Presence-only, relative abundance, species richness, non-EBV genetic analysis.
	Spatial range	High: >15.000 km2.	M
		Moderate: 5.000 to 15.000 km2.
		Low: <5.000 km2.
	Temporal range	High: >10 years.	M
		Moderate: 2 to 10 years.
		Low: <1 to 2 years.
Assigned relevance by main classifiers (criteria: majority vote)			M
Modulators	Multispecies	>10 species.	FALSE
	Threatened species	Assigned a threatened level by IUCN.	TRUE
	New species to science	A first description of a species.	FALSE
	New regional species	A first record of a species in the region.	FALSE
	Priority area*	Data from a geographic area of special interest.	FALSE
Assigned relevance by Main Classifiers + Modulators (criteria: if any TRUE, relevance by Main Classifiers increases one category)			H

Note:

^*For our study case, the ATLAS database, we prioritize areas from Northern Quebec.

First, the main classifiers determine whether the dataset relevance is “High”, “Moderate”, “Low”, or “Non-relevant” based on a majority vote (see Table S2 in case of disagreement), and then Modulators increase one relevance category if any of them is present (e.g., “Low” to “Moderate” but not “Low” to “High”). Modulators do not affect datasets classified as “Non-relevant” by the Main Classifiers.