Wednesday, July 15, 2015

Data Science in the Newsroom: Embedly/Washington Post Hackathon creates wonderful tools




The third instalment of hackingjournalism.com event was held at the Washington Post with over 100 journalists, technologists and students in attendance to create some really cool tools over two days. Looking at data science - what it means and how it can help journalism, the creatively fun event across one weekend, organized by Embedly and MIT’s Future of News initiative, among other sponsors, saw 13 teams and 13 projects, created in about 30 hours.

The projects produced some amazing tools from a kind that helps journalists by making their bookmarked articles easy to file and search to another that reminds people about what they read a year ago ; offering suggestions for catching up with many other ideas churned out.

The hugely successful HackingJournalism Data Science hackathon had participating teams of between three to six people - developers, journalists, designers, and students who had never known each other before the start of the event but formed exceptional bonds to create great projects.

Below is a brief on the various tools created by the teams:
  • ·         Quotable allows people to easily highlight and share among friends small slices of information they find in online stories.
  • ·         FileThis looks at a problem that plagues not only journalists but anyone who is doing digital research — how to keep track of all the bookmarked information you have found on the web. This tool allows you to easily file, store, and search the material, as well as add tags.
  • ·         Audiolyzer takes bar charts and turns them into musical tones. (It was a crowd favorite. As was their choice of data — shark attacks. Do remember: It is “Shark Week” time.) It seen as being useful for the visually impaired, data scientists, and maybe in podcasts.

  • ·         Journofriend. A common problem for journalists is finding a fresh angle on a story that has been heavily covered by many news organizations over a period of time, such as the Ebola outbreak. Their tool helps reporters find new sources on the topic.
  • ·         Dig.It is a real-time traffic monitor that aids in newsroom decision making. It decodes Twitter traffic to focus on what people are talking about and helps news organizations track trends and surface related archived material that can be repurposed.
  • ·         The Bridge attempts to… well, bridge a big divide between what readers are talking about and what reporters are writing, based on tweets and hastags.
  • ·         Then there’s Flindr (as in a mashup of Flipboard and … Tindr). How much time to do you have to read an article? Check off on the dashboard if you have 1 minute, 5 minutes, 10 or longer. And then stories are recommended for you. The system also adjusts story lengths as it learns how fast you read.
  • ·         Team Decision Fatigue has an interesting take for its tool, DCSN. It tries to find the users with the best profiles to show calls for action and then serves them up buttons personalized for their use.
  • ·         Make Me Smarttries to help a constituency that probably does not get a lot of love — the press secretaries of politicians. But the poor blokes struggle to make sense of a hurricane of information that is overwhelming. So MMS tries to take a “noisy world and make it as neat as an organized desk,” as one member put it. It helps them find geo-located information and experts from within their state that can help them out.
  • ·         Renaissance Broccoli tackles a problem that puzzles a lot of people — why do some high-quality, fascinating stories fail to achieve traction? Some stories you would bet would go viral, or at least attract a major audience, but fail to take off. Their tool surfaces those “hidden gems” by looking for stories with four times as many comments as reads, and gives them a second push.
  • ·         Continental Magical System is a play on CMS, the acronym most reporters and editors love to hate. Basically because the CMS they use almost invariably sucks because it is so complicated and non-intuitive. This system would allow reporters and editors to work together on the CMS, but using email. It’s a simplified, streamlined system.
  • ·         Project full_storyis designed to broaden your horizons so you are not stuck in an echo chamber. It looks at what you are reading and finds articles on similar topics but written with dissimilar material, and then presents you with the different takes.
  • ·         oldNEWS is a “throwback recommendation system.”It finds the articles you were tweeting and sharing in the past year or years and offers them up to you, complete with new recommendations on current stories for the old topics.



No comments: