“…Other authors highlight that estimating the difficulty of annotators could improve the reliability of the annotator performance estimation, which could lead to more powerful models (Aung & Whitehill, ; Cao et al, ; Duan et al, ; Nguyen‐Dinh et al, ; Ni et al, ; Wan & Aggarwal, ). There are some classic crowdsourcing algorithms that take into account instance difficulty, such as (Donmez, Carbonell, & Schneider, ; Welinder, Branson, Perona, & Belongie, ; Whitehill, fan Wu, Bergsma, Movellan, & Ruvolo, ), which could be used directly or as a basis for the development of new algorithms that consider the restrictions of the problem at hand, such as the scalability to a large number of examples or the time complexity of the algorithm (Rodrigo, Aledo, & Gámez, ).…”