Only slightly though. It hardly seems practical to try and infer gender from names, in a way where it can’t be obtained through historical records, or the user.
No, since the dataset is bound to give nonsensical results, we search for sources that are more precise. Hint: “Andrea” already mentioned and Japanese names
Only slightly though. It hardly seems practical to try and infer gender from names, in a way where it can’t be obtained through historical records, or the user.
For a given individual, sure. If you’re trying to do some statistics over a whole group that you have no other record for, it could be useful.
Sounds like those statistics output would the heavily biased by whatever process you were using to turn names into genders. In short, a bad idea.
“Since the dataset isn’t 100% perfectly annotated for analysis, we should give up the whole project entirely.”
No, since the dataset is bound to give nonsensical results, we search for sources that are more precise. Hint: “Andrea” already mentioned and Japanese names