This week’s list of data news highlights covers April 19-25 and includes articles on the new data journalism initiative at the New York Times and an initiative to predict health changes at a children’s hospital.
The CEO of student data consolidation nonprofit InBloom announced that the company would fold this week. The Gates Foundation-funded effort would have worked with school districts to securely store student data in a single repository to aid in analysis and promote the creation of personalized learning products. InBloom was ultimately unable to overcome backlash from parent groups who felt that its data security measures would be insufficient to protect students’ privacy, despite commitments to follow industry best practices for data stewardship.
This week, NASA announced that its Kepler Space Telescope had discovered an “Earth-like” planet about 500 light-years away. The achievement is a case study in data reuse. The telescope malfunctioned in 2013, limiting collection of new data, but scientists were able to mine an archive of its older readings to find evidence of the planet. Two full years of data from the telescope have not yet been explored.
The Boston Children’s Hospital has joined with data science startup Etiometry to develop predictive analytics for forecasting changes in patient health. Using data from a wide variety of sensor-based sources, such as heart and breathing monitors, the technology models patients’ risks of getting worse. The creators hope their technology will help take some of the guesswork out of evaluating a patient’s overall health, offering doctors a simple numerical score to use in deciding how to treat a patient.
The New York Times launched “The Upshot” this week, a news vertical that integrates data journalism and explanatory reporting to give readers an intuitive understanding of the quantitative side of complex stories. The site joins recently launched data journalism outfits Vox and FiveThirtyEight in a wave of efforts to use data visualization and analysis to aid reporting. The Upshot will also feature deep dives into data related to other, traditionally-reported stories from the New York Times.
The Good Judgment Project, an academically-led effort to crowdsource predictions about world events, has established a track record of success over the past several years and is currently recruiting a new round of citizen predictors. The project, sponsored by the Office of the U.S. Director of National Intelligence, has addressed 250 questions so far and has consistently performed better than individual experts. The projects organizers have chalked their success up to the fact that that ordinary citizens participating in the project do not have other incentives besides getting the correct answer, while individuals in organizations may find certain predictions politically untenable.
The State of Kentucky’s data sharing and analysis initiative, launched last year to provide better care to frequent emergency room patients, is beginning to bear fruit. The program has identified 4,400 Medicaid recipients as “super-users” of emergency room care. Hospitals can access data from the network to view these patients’ histories and offer personalized counseling and treatment. While it is still too early to see improvements in outcomes, the organizers hope the system will allow doctors and hospital staff to direct patients to primary care providers and other, more appropriate settings.
The Food Network has bought Food on the Table, an Austin-based startup that uses analytics to help consumers plan meals according. The startup works with local grocery stores to track sales and then recommends shopping items based on these deals. The Food Network, which houses a large database of recipes on its website, could use Food on the Table’s service to generate meal plans based on its recipes.
New York-based nonprofit SumAll.org is applying data science for social good across a range of fields, from tracking global human trafficking to combatting homelessness in New York City. The organization, an offshoot of commercial data firm SumAll.com, is also working on a project to forecast prescription drug abuse risk and documenting casualties of the Syrian crisis. The company focuses on data-rich, high-impact projects and works to leverage already-existing data to generate insights and drive decision-making.
Bay Area startup BaseHealth launched this week with a mission to deliver health recommendations to patients based on genomic data, medical research, and Internet of Things data. Its platform can identify risks of more than 40 common diseases from genetic analysis and guide patients toward healthy lifestyles to mitigate those risks. The information will only be available through a doctor’s order for now, in order to get around many states’ medical data access laws.
Microsoft is calling on the public to help refine its Cortana personal assistant product. The technology is based on machine learning algorithms—algorithms in which performance improves with the addition of more data—and is currently described as an in-progress beta release. Cortana’s “personality” will be designed to encourage more back-and-forth with users, in hopes of collecting as much conversational data as possible.