Gossip Machine uses a years’ worth of page views to surface potential “news days” in that year for any topic with a Wikipedia page.
It does this by generating a daily average of page views and then flagging dates which are between 150% and 190% of the average (depending on the setting you use). It then generates pre-filled Google News and Google Web search links for those dates so you can go straight to a (hopefully-useful) Google search for your topic.
Try the name of your governor, your favorite musician, your favorite brand of soda, the closest metro area to your hometown, etc.
Gossip Machine works best with pages that have an average of at least 7000 page views a day. Lower than that and results might get wonky.
Want to learn more about how Gossip Machine works? Check out this ResearchBuzz article!