Hungarian researcher Csaba Molnár from Eötvös Loránd University has developed software in an attempt to determine what a dog's bark means. According to Telegraph.co.uk "The software has learned the nuances of woofs, howls, yaps, snarls and growls in various situations and is now able to classify dog barks with reasonable accuracy, along with the identity of the animals themselves." Molnár says "If we could find the acoustic characteristics of barks which reflect to certain emotional states of dogs we could gain information about the dogs' "well-being" which would have several applications on the animal welfare field." but don't get too excited, the accuracy is only 43% in detecting six different states: 'stranger', 'fight', 'walk', 'alone', 'ball' and 'play' and the research has only been conducted on one breed. No word on when they'll try to figure out cats!
