There were 40k twitter accounts identified by the twitter user eclipse_ow before all went to shit about 6+ months ago. These lists were shared with 0_hour1, broken into two lists of 20k
This was the initial twittergate with relation to pedo accounts. Eclipse had their account shut down for distributing the the first 20k. Eclipse managed to use an algorithm to identify these accounts is was all done within about one week. The first 20k they were sure about 100% bad shit going on and the other 20k Eclipse was not convinced enough at that time with the limited resources at hand to publish the second list.
Most of the first 20k are still active and all of the 2nd 20k are active. These lists and the processes used to identify them automatically still exist, the procedures used can be mapped onto other social sites but were not at that time.
I have tried in vain since to find out how the hell they managed this, but I presume it requires scraping the accounts and therefore implicating yourself as now hosting the content. As such they wont share the process externally.
If anyone has contacts with 0 or Eclipse please try and get the information out, they are sitting on a lot more data than what was initially shared, perhaps this is some of that data coming to light now.
view the rest of the comments →
LostandFound ago
There were 40k twitter accounts identified by the twitter user eclipse_ow before all went to shit about 6+ months ago. These lists were shared with 0_hour1, broken into two lists of 20k
This was the initial twittergate with relation to pedo accounts. Eclipse had their account shut down for distributing the the first 20k. Eclipse managed to use an algorithm to identify these accounts is was all done within about one week. The first 20k they were sure about 100% bad shit going on and the other 20k Eclipse was not convinced enough at that time with the limited resources at hand to publish the second list.
Most of the first 20k are still active and all of the 2nd 20k are active. These lists and the processes used to identify them automatically still exist, the procedures used can be mapped onto other social sites but were not at that time.
I have tried in vain since to find out how the hell they managed this, but I presume it requires scraping the accounts and therefore implicating yourself as now hosting the content. As such they wont share the process externally.
If anyone has contacts with 0 or Eclipse please try and get the information out, they are sitting on a lot more data than what was initially shared, perhaps this is some of that data coming to light now.