Test pulling twitter data / mob detection

[AlexanderLill]

Hello everyone,

as suggested on the last day of presentation week we should probably test the mob detection.

I am not sure what the current status on this topic is, but as I said on that day it would be cool to gather all folks with twitter accounts and test if this actually works :)

Lets create a list here what we want to test, and what we need to prepare before that, and then find a date for that soonish ;)

I'll start, everybody feel free to edit this list:

• Test if getting all tweets from twitter actually works (for sentiment analysis)
• Test if getting all tweets with geolocation actually works (for mob detection)
• Test if mob detection works
• Test if handling of twitter-native retweets works

Requirements:

• Working code
• Adjusted thresholds for mob detection (e.g. if 3 people at the same location tweet)
• Adjusted thresholds for mob detection (e.g. radius or sth of 10km ... depending on possible cluster configuration, lets talk about that)
• Ask on google groups who has a twitter account and setup a doodle for a date

Let me know if there is any feedback or if I am completely off with this issue ;)

[sacdallago]

@AlexanderLill very well thought through and this can work. Waiting for results. Anyone?