10 Bits: The Data News Hot List
This week’s list of data news highlights covers January 11-17 and includes articles on an application of Google Earth to monitor global fishing and $2 million in grants for health data apps.
Australian scientists hoping to gain insight into the global die-off of bees are planning to collect data on the insects’ movements using thousands of tiny sensors. The sensors, which the scientists will attach to the backs of worker bees in Tasmania, use radio frequency identification technology to track each animal’s location. Since exposure to pesticides and other environmental changes can affect bees’ behavior, the scientists hope location-tracking will help pinpoint when and how bee populations begin dying out.
Sociologists at UCLA have developed statistical predictors to determine which films are likely to get an Academy Award nomination. The research, which will be published in February in the American Sociological Review, looked at nearly 3,000 films from between 1985 and 2009. Significant predictors of Oscar nomination included genre—dramas and war films performed better than horror flicks and action movies—as well as the presence of a previously Oscar-nominated director.
A recent study from the University of British Columbia used Google Earth to remotely monitor fishing in the Persian Gulf, finding that the number of catches in the fragile ecosystem may be six times greater than the numbers officially reported to the UN’s Food and Agriculture Organization. The researchers looked for intertidal fishing weirs, large structures that collect fish with the ebb and flow of the tides. Experts at the World Fish Research Center in Egypt hope the technique will be applied in the future to monitoring and enforcing international fishing agreements.
The Philippines launched its first national open data portal this week with data.gov.ph. The site consolidates data from 19 government agencies, and features an applications page with various tools that help users explore and work with the data. While still a work in progress, the site already offers around 650 data files.
BlueLabs, a company formed by former members of President Obama’s data squad, is tracking the treatment patterns of people who enroll in new insurance under the Affordable Care Act. The team gathers information on variables, such as how frequently the newly insured visit emergency rooms, and compares that with previous behavior.
As more government agencies move to the cloud, large-scale data analytics is increasingly within reach at low costs. Raleigh, North Carolina’s Wake County Emergency Medical Services, for example, is using the cloud to store and analyze data on cardiac arrest outcomes, recently using the results of their analysis to change its recommendations for how long to perform CPR on cardiac arrest patients. Previously, recommendations had only called for 25 minutes of CPR, but the new analysis found that CPR could continue to save lives for up to an hour.
Scientists at Stanford University and the University of Washington have demonstrated a data-driven method to identify the fish present in a 1.2 million gallon body of water from a sample the size of a single glass. The method conducts DNA analysis of fish skin cells present in the water and reports the relative abundance of various species present in the sample. The scientists hope the method will be applied in the future to conducting a census of fish in the open ocean.
Researchers from three UK business schools have developed a new approach for retail companies to improve customer delivery services and reduce costs with the use of predictive analytics. The approach uses customers’ previous orders to predict when they will likely request deliveries and create model schedules based on those predictions. The researchers ran a simulation of this model on a major UK e-grocer and demonstrated a 4 percent increase in average profits over six months.
This month, the Texas Advanced Computing Center, HP and NVIDIA launched Maverick, a visualization platform for high performance computing applications in the science and engineering communities. The system is designed to allow researchers to explore large data sets using interactive visual analysis techniques. Scientists have numerous resources for large-scale computing and data analysis, but visualizations often have to be created on smaller datasets using much less granular data.
The Knight News Challenge: Health concluded this week with the announcement of seven winners and more than $2 million in grants. The Knight Foundation challenged technologists to create data-driven solutions to address health issues such as prescription drug abuse. The winning projects include a real-time dashboard with information on health care enrollment, cost, and outcomes metrics in the Camden, New Jersey metro area, as well a set of low-cost chemical analysis tools to allow individuals to track hazardous chemicals in their local environment.
Photo: Flickr user Mighty Boy Brian