Please expose scaling coefficients on task ranking.
 You can sort a task list using graph network methods with or without geometric embeddings of those graphs, and not just in a small number of dimensions. Every semantic distinction like one/many, male/female, food/nonfood, frequency of occurrence, etc, are all potential dimensions to use to find the most important missing pieces. When you do, how do you scale the relative numeric values of the dimensions?

Please allow us to see the numeric scaling functions in task sorting views.

P.S. @GoranSMilovanovic, Please post the Wikimedia DE links for inter-project Wikipedia completeness. I copied them to the chat at the WikiEdu meeting half an hour ago, but forgot to save them. Do they have a multi-language version of https://en.wikipedia.org/wiki/User:DataflowBot/output/Popular_low_quality_articles_(id-2) ? It looks like https://wikirank.net is one, as per https://www.mdpi.com/2073-431X/8/3/60/htm

I want to scale task lists whether the sort order determinants are categories or numeric values, but I'm not exactly sure how to do it when the sort order has only one determinant as in https://github.com/wikimedia/analytics-wmde-WiktionaryCognateDashboard/blob/master/Wiktionary_CognateDashboard_UpdateProduction.R#L389 but I think this is easily doable with the multiple numeric determinants at https://github.com/wikimedia/analytics-wmde-WiktionaryCognateDashboard/blob/master/Wiktionary_CognateDashboard_UpdateProduction.R#L207 assuming -- and this is a big assumption -- that the rows and columns are comprehensible to users who might adjust what starts as a list of [ 1.0] text fields. (Or are they more like [ +0.0] ? https://en.wikipedia.org/wiki/Bhattacharyya_distance )

The requests presented in T202610#5535326 and T202610#5535427 are unclear.
Please define the requests in a clear and concise language (I am discouraging the use of mathematical or technical language, but it must be made understandable and refer directly to the data products offered at the Wiktionary Cognate dashboard), and then we can consider them.