@EvoEcoEric Here is a comparison study, Eric - FWIW the above article does acknowledge some (but def not all) the problems with the methodology https://t.co/jnB8CL3OLs
@HarmitMalik Just in case.
The authors used https://t.co/D9E7v3dwts to identify the gender of the authors. But with all these kind of API, many persons remain misclassified. See for example https://t.co/rG4gtWevzn https://t.co/eD7lailCYN
Oh, this is nice: This article of @h_mihaljevic and myself is one of the top 5 most viewed #Databases articles published in @PeerJCompSci (now Web of Science listed) https://t.co/kaWklqEWuU
@schock @genderify 1 / There are a lot of services like this out there. They may be better trained. But the concept is almost the same. I think there are circumstances in which you simply do not have gender info and you would benefit from having even approximated info. https://t.co/Jmi8E26WjY
@Kevbonham Finally, if integrating within our codebase is cumbersome, you could just genderize our labeled test data https://t.co/oF883E4MUU and compare with the results from our 5 evaluated services. The metrics we analyzed are described in our paper https://t.co/Ww6SNBnJ1B @h_mihaljevic
@lusantala Just got around to reading your benchmarks of name->gender parsers, very cool! https://t.co/Zljn27YNrt Is it easy to add another thing to benchmark? I started developing something for julia, and it would be nice to see how it stacks up https://t.co/BCnIbL5Htf
@LangSciPress What did you use it for? How was your experience with it? Thinking a lot about gender guessing (how to do it/if one should do it at all) lately. See this overview by @lusantala & @h_mihaljevic
https://t.co/uclVvPXgtG
Our article 'Comparison and benchmark of name-to-gender inference services' was one of the top 5 most viewed #Databases and #DigitalLibraries articles published in @thePeerJ journal in 2018!
When analyzing #gender in #academia - how to determine the gender of researchers? A critical analysis can be found here: Santamaría L, Mihaljević H. (2018) Comparison and benchmark of name-to-gender inference services. PeerJ Computer Science 4:e156 https://t.co/sb7w4m2tTE #oaweek