Once again, big data comes to the rescue ... for identifying birds.
A new app, Merlin, published by the Cornell Lab of Ornithology uses a database of 70 million bird sightings from the citizen-science eBird project. This data is used to determine what bird species are likely to be in your area.
Next, based on insights into how people see and describe birds ("By participating in online activities to describe birds based on photos, [bird watchers] contributed more than three million data points that Merlin uses to deduce which birds are viewed based on people's description of color, size, and behavior"), you are asked five questions about the bird's size, it's main colors, where it's feeding, etc. and voila! Your sighting is identified.
A selection of Merlin app questions to identofy your bird sighting
To improve accuracy user feedback is used so that mistakes and unexpected sightings are incorporated into the data providing a source of constant improvement.
The app also displays professional photos of birds (some 1,400 are currently available) along with descriptive text, sound samples, and range maps.
This is a brilliant idea and beautifully executed. And as an example of capitalizing on data resources and turning them into effective, useful information it's outstanding.