Can We Get IP of Tweet?

[pmacMaps]

Is there anyway we can add an IP address or other data piece that we could create a map from? It would be interesting to create a web map showing where these are coming from?

If it's not possible, then just close this issue.

[rzj]

Probably not possible. Public twitter api does not contain IP field.

[shintakezou]

It would interesting to know how they decided that these handles belong to "IRA". At first I thought they had access to all the data Twitter could provide. In this case, at least they can say "this user twitted from this IP which is a russian IP", or something like that. Without this kind of info, these are just tweets of users who aren't aligned to the political discourse of parties who've lost the elections (or in general, power over, and trust of voters), selected because of this and not because there's evidence of anything. If this is the case, it would be very sad.

Add. It seems the list of handles was given by Twitter, but still we don't know the selection criteria; the README says

The researchers also removed 19 handles that remained on the June 2018 list but that they deemed very unlikely to be IRA trolls.

So... these are the known false positive.

[pmacMaps]

@shintakezou, I only saw this repo because it was listed as GitHub trending repo. It would indeed be unfortunate if the process used to identify these tweets as being connected to IRA was suspect.

[shintakezou]

It would be interesting to know how, though surely the details can't be released. It would be even more interesting if there would be some kind of recurring pattern, like a signature, which ties these handles and that can be seen in this public data set.

About discrediting the data set (comment in my mailbox, here I can't see it): yes, I think there's 50% smoke and that the matter is more a political (and geopolitical) one than a real security threat or an actual attack on democracy. Otherwise, we all should be very worried because it would mean we live in really fragile democracies populated mainly by brainless puppets in stead of mostly aware citizens.

Now that some of this “Russian fever” has surfaced in the Italian first pages of several news papers using mostly this very same data set, plus other statements we can't check (no public data released — we are all believers, aren't we?), I am even more convinced that the matter is more about internal politics than concrete concerns for what Russians do using Twitter or other social networks.

Mod:

This PDF document explains the criteria.