Network analysis of alt-right and militia groups in online communities

[gati]

Which nodes of a given community are most influential?

How does information or sentiment spread within communities?

When did a given community emerge? Did it emerge from another community? Is it expanding or contracting?

[gvdr]

I think this is a very important set of questions.

To be ambitious, we may try to go predictive, such as:

• providing a warning system for far right bots and militants;
• detecting nascent memes within far right communities before they explode;
• ...

[gati]

Absolutely agree - re bots, this might be an interesting jumping off point! http://truthy.indiana.edu/botornot/

[gvdr]

Didn't know that paper. It's on my reading list now.

[mollyegibson]

Hey guys! I'm new to d4d and came across this issue while exploring the github--and I'm really interested in these ideas. I'm not quite sure how to get started and would love some guidance! I'll give you a bit about my background: I'm currently a master's student in engineering data analytics & statistics. I'm relatively new to CS, but I feel pretty confident with Python, R and SQL and I've done a few machine learning/data science projects (both for class and from Kaggle). I have a lot to learn, but I'm totally geeked on data and I'm excited about the opportunity to contribute to a meaningful+relevant project! Let me know how I can help!

[sjacks26]

Hey @mollyegibson! Check out #discursive-commdetect in Slack. I think gvdr and henripal have been thinking about how to get started, I'm sure they can use your help!