More hairballs?

(Ok, some of the graphs are not hairballs, but seriously, what can be inferred from looking at the NFL graphs?)

The NFL graphs are not graphS. It is one graph. In the NFL case, each graph is a snapshot corresponding to monthly activity.

completely needless distinction, because subgraphs are graphs

Signal processing on graphs is amazing, could be a way to implement reasoning in AI.

I don't find this particularly compelling without seeing the paper where they discuss the technical details.

They also noticeably fail to link their previous paper while having lots of links to Wikipedia about topics.

Link to the paper: https://drive.google.com/file/d/0B38Is3bZ7gigVUtfTW9nTGc3T2c...