We at Inform believe that data tells a story across all industries, and so every week we round up the most interesting ones right here. The latest: the Internet of farm things; helping patients with epilepsy; earthquake?
The agriculture industry is no stranger to technology. Since the 1990s, farmers have collected data via GPS and USB ports on farm equipment. The issue was taking the time to download and analyze that data.
Thanks to John Deere, the Internet of Things now includes tractors and other farm equipment, and thanks to a data collection firm, precious data will be available via the cloud. Farmers can use such data in a practice called precision agriculture, in which information collected by GPS and sensors can be analyzed “to find better ways to irrigate, discover new seed varieties or target areas of their acreage that need more fertilizer.”
In other words, farmers will be able to more easily use big data to help make “growing crops more efficient and productive.”
Using a big data approach to customer and campaign management, T-Mobile has succeeded in a 50% reduction in churn rate, that is, the annual percentage rate that customers stop service.
The telecomm industry has access to “thousands of data points across millions of customers,” but only T-Mobile has leveraged such data to achieve measurable success, staying abreast of “trends by region, customer service inquiry patterns, purchases by location and customer lifetime value.”
From the data, T-Mobile understands which customers are potential influencers, and which might not be having a “top of shelf experience,” and to be able to remedy those situations sooner rather than later.
Based on a “dataset of speeches and debates in the UK’s House of Commons in the years from 1975-2015,” researchers in Europe and the U.S. are looking to determine “trends and indicators for political cohesion.”
This is only “an initial foray,” but it’s already revealing interesting patterns, such as “the fact that both Labour and Conservative speakers have a historical tendency” to promote topics they, and no one else, are interested in, and “that topics tended to stop clustering during periods of political certainty.”
The analysis also shows the rise and fall of certain topics. For example, health care “decreased dramatically throughout the 1980s and early 1990s,” while welfare rose to take its place, and education fluctuates but otherwise remains consistent for the entirety of the 40 years of the data set.
Using an open source framework from Apple, researchers at Johns Hopkins have created an Apple Watch app “to collect data from patients with epilepsy before, during, and after their seizures.”
The purpose of collecting data such as “physiological changes, altered responsiveness, and other characteristics of recurrent seizures,” is to gain a better understanding of the neurological disorder, and “to develop new methods for (and determine the role of technology in) monitoring and managing the disorder.”
Doctors often ask patients with epilepsy to record their seizures. But because patients often lose consciousness during a seizure, this can be a challenge. The Johns Hopkins app would automatically capture that episode and other useful data as well.
While the US Geological Survey (USGS) National Earthquake Information Center has access to about 2,000 earthquake sensors, those sensors are mostly in the US. As a result, much of the rest of the world isn’t covered. To compensate, the USGS has started using Twitter data.
Using Twitter’s public API, USGS was able to apply filters to shake out earthquake-related tweets, finding “that users tweeting about real earthquakes tend to keep their posts short — seven words or fewer,” and that “users who are actually experiencing earthquakes don’t includes links or details about magnitude in their tweets.”
USGS discovered that Twitter data can indeed be “an effective way of detecting earthquakes, typically delivering alerts in under two minutes.” A year ago they were able to use tweets to detect an earthquake in Napa, California in less than 30 seconds.