On 11/9/23 21:09, 4st nomic via agora-business wrote: > These are based on the email used to register according to the registrar's > report. > My plan here is to omit them entirely from the dataset, since they are most > definitely bad data. > If I can find evidence on whether or not they are the same person either > way on any of those players, then I can include them again.
Ohhh, there's the missing piece, you're working based off email similarities. Knowing that, I think we can narrow more, and ask the Registrar to work on consolidation. Most of them have the same email, and I would argue are pretty clearly the same person. Some of them have similar emails, and while they look likely to be the same person it's not conclusive from this evidence alone. IMO, consolidate the ones with the same email into 1 person for your analysis, exclude the others. -- nix
