Abstract
Most of life’s diversity lies within “Dark Taxa” that are abundant, hyperdiverse and difficult to identify. Coupled with this challenge is the very limited taxonomic capacity and expertise available to study these taxa at species level. Phoridae is an example, with ~4700 known spp. and this likely to be only 10% of the World’s phorid fauna. Phorids are often caught in large quantities by Malaise traps. Our approach to identifying large numbers of phorids links historical and modern data with machine learning.