10 Bits: The Data News Hot List
This week’s list of data news highlights covers November 9-15 and includes articles on an athletic clothing startup’s push into wearable computing and a new set of U.S. public-private partnerships around data.
The White House announced a variety of new public-private partnerships around data this week. One is CancerLinq, a five-year, $80 million partnership with the American Society of Clinical Oncology to consolidate cancer data sources and provide real-time decision support to oncologists. Another partnership, between NASA and Amazon Web Services, will make NASA’s Earth Exchange information-sharing network available to the public for free through the Amazon cloud.
This week, Stephen Wolfram announced the creation of Wolfram Language, a new high-level programming language that was derived in part by the software underpinning Wolfram’s popular Mathematica software. Wolfram, the computer scientist and entrepreneur who also created the Wolfram Alpha computational knowledge engine, calls the new language his most important project.
In developing countries, government-led open data initiatives are often difficult to get off the ground due to lack of funding and resistance to transparency. In some cases, non-profits have stepped in to accelerate open data efforts, such as the Parliamentary Monitoring Group, which helps voters learn about local politicians in South Africa, and Odekro, a Ghanaian website that monitors politicians’ behavior and provides online access to certain public records.
U.S. sports clothing company Under Armour announced the purchase of fitness app company MapMyFitness this week. Under Armour’s CEO says the acquisition will help his company move into the wearable tech and quantified self space. Earlier this year, the company released Armour38, a chest band that tracks heart rate and workout intensity, among other health variables.
Massively open online education platform Udacity announced a partnership with scalable data software company Cloudera to provide a free data science curriculum online. The first offering from the program will be an introduction to Hadoop and MapReduce. Cloudera currently offers a range of in-person trainings and certificate programs, but none in a widely available and free format.
MIT researchers have developed a new technique to help automatically deduce the structure of certain kinds of data, which can significantly speed up modeling and analysis. The group applied their technique to a sample data set of commercial airline flights and were able to efficiently infer information about how flight delays propagated in that particular airport.
New York City’s education department revealed several privately developed online and mobile applications that help students navigate the complicated high school choice process and run on school-level open data. The apps were created as part of the city’s School Choice Design Challenge and included a decision support system that helps students rank their educational priorities and a recommendation engine that suggests schools based on the choices of similar students.
New research published in the Proceedings of the National Academy of Sciences calls into question the statistical significance levels scientists use to argue the validity of their results. The author of the research, a professor of statistics at Texas A&M University, argues that scientists should set significance levels to be ten times stricter than they currently are. The research is particularly timely due to other recent papers alleging that many results, particularly in the biomedical sciences, are not reproducible and may arise from insufficient evidence.
Amazon announced stream-processing service Kinesis this week, joining the Apache Software Foundation’s Storm and IBM’s InfoSphere in the market for technologies that can handle data “firehoses,” such as the real-time Twitter feed. Amazon’s managed solution does not require clients to invest in any infrastructure, instead they use the cloud-based data storage and retrieval technologies that Amazon has already developed.
Car insurers, like Progressive in the United States, Tesco Bank in the United Kingdom and Generali Group in Italy are hoping consumers will install monitoring devices in their cars in exchange for lower rates. The devices, like Progressive’s Snapshot, monitor speed, time of day and braking intensity to set premiums according to how safely an individual drives. At present, however, use-based policies are only available to 2 percent of the U.S. car insurance market.