At the “H/H @Stanford: Computational journalism with CIR, Vocativ and SmartNews” meet-up on Tuesday night, journalists (“hacks”) and programmers (“hackers”) gathered to learn from three speakers about ways to use technology to enhance news narratives. Computational journalism uses data to find interesting trends to generate stories and help complement them, such as through graphics.
The Stanford Computational Journalism Lab joined with the organization Hacks/Hackers to sponsor the event.
“The idea was to bring together journalists and technologists to talk about the future of storytelling,” said Burt Herman ’95 M.A. ’96, the founder of Hacks/Hackers.
The first presenter, Gerald Rich, interactive producer at Vocativ, discussed the usage and effectiveness of using maps to illustrate news stories. To him, maps are tools for data visualization.
“This is pointing to where it is and giving the reader more context,” explained Rich, as he pointed to locator maps of Tianjin, China and Palmyra City, Syria. “As the world becomes more and more connected, we are hearing more stories about these places and often readers don’t know where they are. This can be as simple as highlighting countries.”
However, maps can be misleading as well, and Rich elaborated on how people have tried to rectify biases, such as through cartograms.
Naoki Orii, a software engineer at SmartNews, unpacked how the SmartNews app chooses news recommendations for its users by using an “exploration” mode — or choosing articles outside of the user’s preferences in order to enlarge their knowledge of the world. This type of model contrasts with the “exploitation” model, which only recommends articles within the user’s preferences.
“For example, you go to a restaurant in downtown Palo Alto. Exploitation is going to your favorite restaurant, whereas exploration is the equivalent of trying a new restaurant,” Orii said.
However, the risk is that the user may not be satisfied with the articles the SmartNews app recommends. Orii ended his presentation by asking the audience on how to achieve a balance between the two models and whether there were news events and articles that everyone should be universally recommended.
The last presenter was Allison McCartney M.A. ’15, a data fellow at the Center for Investigative Reporting (CIR) and Magic Grant Recipient at the Brown Institute for Media Innovation. She described her work mining the data from the National Missing and Unidentified Persons System (NamUs) to launch a more user-friendly website to help solve cold cases.
“Part of the problem is shoddy police work — people not taking DNA samples, but part of it is a stupid technological question, and it comes down to a website,” McCartney said.
The current NamUs website separates the Missing Persons and the Unidentified Persons databases, even though the combination of the two may be helpful to identifying the corpses in the morgues. The new website by the CIR combines the two databases and includes photos of the missing persons and corpses to help the amateur sleuth community better solve these cold case files of missing and unidentified persons.
So far, the community has identified 100 matches between the missing and unidentified people, but McCartney speculated that only a third of those matches are real. However, there is one match that seems especially promising, and she hopes that they will be able to solve this cold case.
“We’re just happy to be able to make an impact on people’s lives,” McCartney.
Contact Anne-Marie Hwang at amhwang ‘at’ stanford.edu.