haha, no, this was an example query in the application… I came up with some unique ones of my own, most of them were too controversial to post :P
To answer your questions, here are my theories:
The second spike, which is around October 28th, is probably Halloween. Since Oct 31st was a wednesday, most folks moved the party dates to the preceding weekend.
Hangovers are more talked about than the party, which will probably have it’s own “event” page. Also, there is only one word for “hangover”; party has many synonyms.
Overall, the spikes are also a tell-tale sign of Facebook usage; far more users log in to the website over the weekends, compared to weekdays. I wish they would provide a normalized graph as well.
haha, no, this was an example query in the application… I came up with some unique ones of my own, most of them were too controversial to post :P
To answer your questions, here are my theories:
Overall, the spikes are also a tell-tale sign of Facebook usage; far more users log in to the website over the weekends, compared to weekdays. I wish they would provide a normalized graph as well.