An innovative artificial intelligence algorithm developed by canine behavior experts is designed to accurately assess the temperament of working dogs to facilitate better training and placement outcomes. This collaboration between academia and canine technology startup Dogvatar uses C-BARQ survey data to predict canine personality types, providing a new approach to human-canine matching. The algorithm generates a canine personality fingerprint, much like the popular Myers-Briggs test does for people.
A multidisciplinary team of researchers specializing in canine behavior and artificial intelligence has developed an AI algorithm that automates the high-stakes process of assessing a potential working dog's personality. They hope to help dog training facilities more quickly and accurately assess which animals have the potential for long-term success in careers such as assisting law enforcement and helping people with disabilities. Personality tests can also be used to match dogs with people and help shelters properly place animals, thereby reducing the number of animals that are returned as unsuitable adoptive homes.
Scientists from the University of East London and the University of Pennsylvania conducted the study on behalf of their sponsor Dogvatar, a canine technology startup based in Miami, Florida. They announced the research results of the dog personality testing algorithm in the paper "Artificial Intelligence Method for Predicting Dog Personality Type" published in "Scientific Reports" on January 29, 2024.
The AI algorithm draws on data from nearly 8,000 responses to the widely used Canine Behavior Assessment and Research Questionnaire (C-BARQ) to train itself. For more than 20 years, the 100-question questionnaire has been the gold standard for evaluating potential working dogs.
"The C-BARQ is highly effective, but many of its questions are also subjective," said co-principal investigator James Serpell, professor emeritus of ethics and animal welfare at the University of Pennsylvania School of Veterinary Medicine. "By clustering data from thousands of surveys, we were able to adjust for the outrageous responses inherent to subjective survey questions in categories such as dog competition and stranger-oriented fear."
The research team's experimental artificial intelligence algorithm works in part by classifying responses to C-BARQ questions into five broad categories, ultimately forming a digital personality fingerprint for a specific dog. These personality types are identified and described based on an analysis of the most influential attributes in five categories, which include: "Excitable/Affective," "Anxious/Fearful," "Aloof/Predatory," "Reactive/Active" and "Calm/Sociable." These final grouped data points include behavioral attributes such as "Excitement when the doorbell rings," "Aggression toward strange dogs that visit your home," and "Chasing or chasing birds when given the opportunity."
Each attribute is assigned a "feature importance" value, which is how much weight the AI algorithm places on that attribute when calculating the dog's personality score.
Dogvatar and its co-researchers plan to further investigate potential applications for their dog personality testing algorithm. "This is a very exciting breakthrough for us," said CEO and "AlphaPack Leader" Piya Pettigrew. "This algorithm could greatly improve the efficiency of the working dog training and placement process and help reduce the number of companion dogs returned to shelters due to mismatches. It's a win-win for both the dogs and the people they serve."
Compiled source: ScitechDaily