See How Wikipedia Topics Are Shaking the News With a Wikipedia Seismograph

See How Wikipedia Topics Are Shaking the News With a Wikipedia Seismograph

As the Web continues to fill with the detritus of information warfare and the infosewage pumped out by numpties who think they’re going to get rich stealing and regurgitating content, I have been forced to rethink my search strategies.

Building complex queries is still important, but there are scenarios where that’s difficult to do. Take news search around public figures, for example. If you want to get some background on Fred Famousguy and Fred Famousguy’s in the news a lot, you can’t create a query that’s both general enough to cover all of Fred’s escapades and specific enough that it will eliminate junk/marginal results.

As long as you’re focusing on the query in this scenario you can only get so far. On the other hand, you can turn the problem around. Instead of working on the query, you can instead look at the space you’re searching. Is there any way you can make it as small and specific as possible, so you’ll get fewer but higher-quality search results?

Sometimes you can via limiting your search to a top-level domain or a set of domains. Sometimes you can restrict your search by language or country and narrow your search space that way. But my favorite way to create tiny search spaces, especially in news search, is by searching just a couple days’ worth of content at a time. Not only does that give you a manageable number of results, but it also organizes your searches chronologically so you can get a timeline sense of Fred Famousguy.

Of course, the million-dollar question is: how do you know WHAT days to search? In the case of public figures and famous people, it’s easy: you check Wikipedia! Wikipedia has pageview logs that go back to 2017. I’ve used Wikipedia page views for tools like Gossip Machine (part of MegaGladys), which analyzes page view counts for a given Wikipedia topic, identify dates with unusually-high page views, and creates date-bounded Google searches. That approach works great when I’m working with the other MegaGladys tools and I’m taking an overall look at a Wikipedia topic, but sometimes I want to see what kind of news coverage the topic is getting over an extended time period. Gossip Machine is designed to create single-day search spans and doesn’t give a comprehensive idea of page view changes over time.

So I made the Wikipedia Pageviews Seismograph ( https://wikitwister.com/ws/ )!

A screenshot of the WPS in action, showing a comparison of Cardi B, Megan the Stallion, and Doja Cat's pageview derivations between June and July of 2024. It's a pretty busy graph.

The idea behind the WPS is measuring the impact of public interest on a Wikipedia topic’s page views. Because public interest means something happened to turn the attention to that topic. Sometimes it’s as simple as being mentioned in a big subreddit like TodayILearned, but more often it’s because there was a news story or event that involved the person.

By visually displaying the deviations from a seven-day moving average in a chart (which looks to me like a seismograph output) you can easily see peaks in the public’s interest in a topic. Of course, that knowledge isn’t very interesting unless you can also discover why the interest has peaked, so the WPS also includes a feature to let you create date-bounded Google News searches using the chart output. Let me show you how it works so you can try it yourself: Wikipedia Pageviews Seismograph is free to use, free of ads, and designed to be used on desktop.

Getting Started With WPS

The WPS getting ready to work. There are spaces for four Wikipedia topics; there is only one being used and that's for Rudy Giuliani. The start date is June 1, 2024 and the end date is August 4, 2024.

The WPS can compare up to four pages at once but we’ll look at that a little later. At the moment let’s take a look at one topic: Rudy Giuliani. That guy’s always in the news for something and it’s hard to keep up. Be sure you’re capitalizing and formatting the topic just as it is in Wikipedia; searching for rudy giuliani will fail.

You also need to provide a start and end date. The start date can be no earlier than January 1, 2017, as Wikipedia did not keep page view logs from the beginning. After you enter a date span and click the Visualize button, you’ll get a graph of page view spikes for that topic over that time period. Let’s look at Rudy’s.

A graph of Rudy Giuliani's Wikipedia page view spikes between June and August 2024. It's very spiky.

Wow, that’s a lot of view spikes. Note that the graph isn’t mapping total page views, but rather deviation from a 7-day moving average. That’s why spikes high above the baseline can cause drops below the baseline; that spike has a big impact on the 7-day moving average. You might also notice that there are no page views visible. To see the page views, run your mouse pointer over the graph. The page views will appear in a tool tip.

Ready to explore some of those spikes? Click on the timeline before and after a spike and two red lines will appear, marking the beginning and end of the time span you want to search.

A GIF showing how to set timespan markers on a WPS graph. A mouse cursor is clicking lines before and after a page view spike.

After you’ve chosen a span of time, the WPS will show you a search form underneath the graph that will take you to a Google News search for the marked span. (You can also choose to remove the search if you picked the wrong dates.)

A gif showing the a time span being marked and a search form for Rudy Giuliani appearing underneath.

The OR or AND options are for when you’re searching and comparing multiple people; we’ll look at that in a moment. For a single-topic search you’re good to go; just click on the Search Google News button. It will take you straight to a Google News search result in a new tab. Let’s see what that big spike is all about.

A GIF showing the full process of marking a timespan on a WPS graph, clicking on the Search Google News button underneath, and going to a Google News search for Rudy Giuliani covering June 30 to July 3.

Ah, that was the span Rudy was disbarred in New York. That makes sense. But is it only the big spikes that can inform us about current events? No; the more famous or prominent the person/topic you’re searching, the more you’ll discover from even small changes in page views. Let’s look at that tiny spike toward the end of Rudy’s chart:

Another span search for Rudy Giuliani, only it's for a much smaller spike in early August. That leads to a story about him settling his bankruptcy case.

As you can see, in this case that small spike also provided some news insight, in this case to the resolution of Giuliani’s bankruptcy case. It only takes a few clicks to set up a date-based search with the WPS, so it doesn’t take long to investigate even little spikes.

The example I just showed you was for one person over a recent time period. But Wikipedia page view data goes back several years, so we can do historical analysis as well for multiple people. Let’s take a look at Donald Trump, Mike Pence, and Rudy Giuliani in January 2021.

The WPS showing a graph for Donald Trump, Mike Pence, and Rudy Giuliani in January 2021. One spike early in the month encompasses all three of them.

In this case you can see there is one time early in that month when all three names spike together, but more often they don’t. When you’re analyzing multiple people and marking time spans to search, you’ll have the option of excluding people from your Google News search and searching with either AND or OR:

WPS marking the dates January 5 - 7, 2021, for a Google News search for Pence, Trump, and Giuliani.

In this case we want to find news stories that involve all three people, so all we have to do is change the radio button selection from OR to AND and click Search Google News. And we find — pretty much what you’d expect to see for that date span.

Running a Trump/Pence/Giuliani search for January 5-7 2021 and getting just about what you'd expect -- lots of stories about certifying elections and various chicanery

But what about that spike later in the month that doesn’t involve Rudy Giuliani? Use the checkbox to exclude him from the Google News search and make sure your radio button is set to AND. You’ll get only search results for Pence and Trump.

Using WPS' Google News search options to exclude Rudy Giuliani from a search for Trump and Pence, January 18 - January 21.

I have been using people for my examples in this article, but are Wikipedia pages of people the only thing that works with the Wikipedia Pageview Seismograph? No! You can do searches like Psilocybin or Pervoskite or Donor-advised fund — any topic with a Wikipedia page is explorable this way. Topics with low page counts may require higher spikes to indicate meaningful results, but they’re easy to find on a topic’s chart.

Back To Top