In the early morning hours of Aug. 25 an earthquake hit northern California, waking up thousands of people. How do we know people were awoken? Jawbone published the data.
Later in the day, the producer of health-tracking wearable devices published data showing that people who wore their UP fitness tracker wristband to analyze their sleep activity suddenly woke up at about 3:20 a.m., just after the quake hit. Virtually all of those living closest to the epicenter -- Berkeley, Napa and Sonoma -- woke up.
Reaction to the data publication has been mixed and is perhaps more interesting than the actual findings of the report. Some commentators say the graphic shows the potential of wearable technologies to influence disaster response and improve public health monitoring. Others have pointed to the obvious privacy implications of gathering and publishing data on the sleep habits of thousands of people.
For its part, Jawbone said the data is anonymized and presented in aggregate, so therefore poses no threat to anyone's privacy.
We can expect to hear more from this debate about data publication in the future.A January report from Accenture found that 52% of consumers are interested in wearable devices, which mainly include fitness trackers. As adoption grows, troves of data will be created. The battle over who gets to analyze, interpret and publish this data has only just begun.
SAS and Hortonworks bring apps to YARN
SAS Institute Inc. and Hortonworks announced a collaboration that will provide better integration between SAS analytic applications and YARN, the main operating system for Hadoop 2. The two companies say this will allow users to co-locate workloads being carried out in SAS application and Hadoop, which will maximize the use of existing resources and eliminate data silos.
Prior to the announcement, it was possible to integrate SAS applications with Hadoop, but it was tedious. Processes that used the same datasets would run on separate servers and require users to make copies of the data. No more, says Paul Kent, vice president of big data at SAS Institute.
"Folks are becoming intolerant of the idea that we should make another copy (and have to reconcile, secure and govern that copy) to facilitate processing," he wrote in a blog on the Hortonworks website. "One of the original themes in Hadoop is to move the work to the data."
The move is part of a larger trend toward consolidating numerous big data jobs onto a single platform. Earlier this summer we heard a lot about how Apache Spark could be that platform.
FDA publishes data hoping for new innovations
Storm is the latest rising star in the Hadoop ecosystem of apps
Learn this definition for YARN