order BTA by interesting-ness - order by (un-harmonized taxonomic names) level of disagreements

[jhpoelen]

when using BTA, I'd like to understand which taxonomic concepts are "interesting", where an interesting taxonomic concept is one in which interpretation varies across taxonomic authorities.

E.g.,

taxonomicId	name	authority
BTA:895@eb557c6d	Neoromicia - see robertsi	name_HMW
BTA:895@eb557c6d	Laephotis - see capensis	name_batnames
BTA:895@eb557c6d	**	name_ALC
BTA:895@eb557c6d	**	name_GBIF
BTA:895@eb557c6d	Laephotis - see capensis	name MDD
BTA:895@eb557c6d	Neoromicia melckorum	name_MSW4

For this example, we see that only MSW and IUCN agree.

We'd like to quantify this degree of agreeness.

Aja's idea is to list all permutations of authority pairs (e.g., MSW vs IUCN, MSW vs HMW) and score them with a 0 if they agree, and 1 if they disagree.

Now, a agreement index can be generated by:

agreeness index = sum (disagreements ) / total comparison

Absolute difference in interpretation across all different authorities.

In our example above, the score / distance (need to find a corresponding existing measure name) would be:

comparison table:

name_HMW vs

[jhpoelen]

see attached screenshot and XLSX for "paper" prototype for agreement index

aja-agreement-index.xlsx

[jhpoelen]

Idea is to compare agreement index curves to compare agreement spectrum across authorities or even within different versions of authorities.

[jhpoelen]

So, we'd like to take the combinations across the name authorities (e.g., name MDD, name_batnames) for each BTA concept.

[jhpoelen]

See also https://jhpoelen.nl/bat-taxonomic-alignment and screenshots below.