This election cycle has seen a proliferation of online tools that make the job of digging up and delivering newsworthy information a little bit easier. The tools–which take advantage of advances in artificial intelligence, algorithms, video software, and analytics–help journalists navigate big data and an internet ecosystem increasingly entwined with the political landscape. The social media wave that helped put President Obama in office in 2008 looks almost quaint in comparison.
CJR rounded up five online tools that will help journalists cut down on research and production time, find new ways of analyzing voters, and create engaging journalism that is digestible for the reader, even without the resources of a big newsroom.
1) Map subcultures on social networks
The morning after the first presidential debate, the hashtag #TrumpWon was trending on Twitter. Gilad Lotan, Chief Data Scientist at start-up studio betaworks, was intrigued by the confused reaction to the trending topic and decided to investigate. Lotan turned to the betaworks platform Scale Model (he is a co-founder), which is used to map communities on social media and track how information filters through interest groups. The tool is geared towards advertisers, but in this instance Lotan used the platform to track the #TrumpWon hashtag back to its source and unearthed a cluster of highly active Trump supporters who were organized enough to Tweet the same message, at the same time, from multiple locations, among tightly-knit networks.
New blog post on the #TrumpWon hashtag and the communities of users who got it to trend worldwide https://t.co/RWmA39vkfr pic.twitter.com/0V0IPvM0Qf
— Gilad Lotan (@gilgul) September 29, 2016
Lotan identified five different categories of Trump supporters who were pushing the hashtag, from traditional conservatives to white nationalists. Later in the day, a fake news story that the #TrumpWon hashtag had originated in Russia began to spread through an equal and opposite world of anti-Trump supporters. Lotan’s Medium post about his research process provides an interesting insight into how densely connected communities from both sides of politics can cause misinformation spread through Twitter and go viral. It’s worth reading the whole thing.
“Trending topics are helpful as they cut across information silos, gaining significant levels of attention from people who would otherwise never see your content,” he writes, going to explain how “false information can spread like wildfire, especially when there are enough people invested in making it true.”
Scale Model offers a 14 day free-trial for those interested in checking it out.
Related: Martha Raddatz and the case for a more assertive debate moderator
2) Keyword-search audio files
After leaked footage of Donald Trump casually advocating sexual assault was published by The Washington Post Friday, it became clear sound bites unearthed from old audio and video material could influence the outcome of the election. Deepgram is a tool that uses artificial intelligence to transcribe large audio files and allow users to search them using keywords. They’re giving journalists free access until Election Day. Try it out here.
3) Make a ‘supercut’
Everyone loves a supercut. After Sunday night’s debate, videos of Trump lurking behind Clinton, Trump arguing with the moderators, and a 15 minute long cut from PBS showing the candidates’ silent moments during the debate, began to do the rounds. Supercuts, which are created by stitching together multiple short video clips, are often used for comedic effect, but they can also highlight trends that aren’t as noticeable during a long broadcast. For instance, this year-old supercut showing CNN commentators and guests uttering Donald Trump’s name 239 times in a 24-hour period encapsulates how Trump can dominate the media cycle and basically sums up coverage of this entire election.
For those wanting to make their own, it’s pretty easy to do thanks to Spin Time TV, an online community dedicated to remixing video of the election. TV News Archive is making a video archive of the debates available on a two-minute delay, enabling the Spin Time TV group to isolate clips, edit them, and share them on Twitter using Mozilla’s Popcorn Maker, a free, web-based, open source video editor. Join the Spin Time TV community and access free tools and tutorials here.
For more political supercuts, such as this one likening Donald Trump to Frank Costanza, head over to The Washington Free Beacon’s YouTube page where you can lose yourself for hours.
4) Visualize the candidates’ campaign messaging
Which candidates goes on the attack more on Twitter–Trump or Clinton? You might be surprised to learn that it’s Clinton. At the time of publication, Trump has posted 2,396 messages classified as attacking, compared to 2,745 from Clinton. The caveat is that Clinton is more likely to post attacks related to issues, while Trump’s attacks focus more on image. That’s the sort of information you can find through Illuminating 2016, a tool that enables journalists to create data visualizations out of the social media activity of political candidates.
The team behind project is compiling data from the Facebook and Twitter accounts of all of the major party presidential candidates and analyzes them in real-time–the website refreshes every hour. Users can make graphs and charts comparing things like the volume of Tweets from each candidate, their most liked and re-Tweeted messages, and their share of promotional hashtag mentions.
For example, Trump’s most re-Tweeted message in the past year was posted on July 20:
The media is spending more time doing a forensic analysis of Melania’s speech than the FBI spent on Hillary’s emails.
— Donald J. Trump (@realDonaldTrump) July 20, 2016
Meanwhile, Clinton’s most re-Tweeted message comes from September 27:
“I never said that.” —Donald Trump, who said that. #debatenight https://t.co/6T8qV2HCbL
— Hillary Clinton (@HillaryClinton) September 27, 2016
Users can create visualizations filtered by time frame, candidate, and topic, generating insights not only into how candidates are communicating with their constituents, but also how their constituents are responding to them. You can read about the methodology here.
Related: FiveThirtyEight’s ‘Whiz Kid’ Harry Enten represents the new generation of political journalist
5) Mine public data
Last month ProPublica and Google News Lab launched the Election DataBot, an online tool that helps journalists find newsworthy leads about the presidential race by automatically surfacing the most interesting details gathered from campaign finance filings, congressional votes, polls, and Google Trends data. Users can search by keyword or browse data by state. The latest data for Hillary Clinton shows a $66,653 donation from Planned Parenthood Advocates of Wisconsin Political Fund, a link to a Politico/Morning Consult poll noting that Trump has managed to maintain support following the unsavory revelations of the last week, and a big spike in Google searches of Clinton’s name around the time of Sunday’s presidential debate.Shelley Hepworth , formerly a CJR Delacorte Fellow, is Technology Editor at The Conversation in Australia. Follow her on Twitter @shelleymiranda.