Hey, I wanted to track the number of mentions that the 3 top candidates for the mexican presidential election of 2018 have during a week, I was gonna use the streaming API with the filter parameter and include their last names: “Anaya”, “Lopez Obrador” and “Meade” like this: stream.filter(track=[“anaya”, “lopez obrador”, “meade”]). Now here’s the thing, people refer to “Lopez Obrador” also as “AMLO”, I thought fine, I’ll just modify that list and instead try this: stream.filter(track=[“anaya”, “lopez obrador”, “amlo”, “meade”]).
This looks fine to me but it got me thinking, so I’d basically be using 1 term to track tweets for Anaya, 1 term for Meade, and 2 terms for Lopez Obrador and that’s where I got a little confused, I know that by including the extra term “amlo” the filter will have one more term to track so of course I’m gonna get more tweets mentioning Lopez Obrador, but since I don’t know how the streaming API obtains tweets I wonder if I won’t, in a way, be “gaming” my own tracker? Like how do I know the filter is not gonna assign more “weight” to the same candidate? I’m filtering a list of terms where 2 terms refer basically to the same thing, I don’t want to make a candidate more popular than he is in reality, anyone know?