#1

Is there any good documentation about how place annotations are applied to tweets?

2 Likes
#2

I figured as much.

1 Like
#3

I revive this post in case someone can help now. Is there any way of getting more information about the subset of tweets that the API provides whenever a places endpoint is used? It is clear that this is a very reduced subset and likely linked to annotations (thus Ed’s question above applies), but I believe it is hard to justify possible selection biases in research without knowing slightly what are the criteria for a tweet to be in.

1 Like
#4

Just a quick curiosity on this. I downloaded twitter activity from a 3 months period in 2017 filtering by location=ES. The place restriction was the only one together with the dates (therefore, it should give me all the activity that attends to those two criteria).

I then used geopy to put coordinates into a random sample of the author.location metadata and here is the result:

map_spain
map_spain1920×1008 93 KB

I know users change this field overtime and it does not need to represent where they live but, still…

1 Like