AI finds five dog personality types

In a groundbreaking study, researchers have harnessed the power of artificial intelligence to unravel the mysteries of canine personalities.

Understanding a dog's disposition is crucial not only for pet owners but also for professionals in working roles. The study, published in the renowned journal Scientific Reports, introduces a novel approach to predicting and categorising dog personalities.

Traditionally, dog personality assessments have relied on questionnaires and statistical analysis. However, this new research takes a leap into the realm of artificial intelligence, a domain well-explored in predicting human personalities.

The team applied machine learning techniques to behavioural data from the University of Pennsylvania’s C-BARQ database which contains over 70,000 behavioural records.

The results revealed five distinct personality types among domestic dogs.

The outcome was five distinct categories, each corresponding to a different personality type: “Excitable/Hyperattached,” “Anxious/Fearful,” “Aloof/Predatory,” “Reactive/Assertive,” and “Calm/Agreeable.”

The research was conducted over two years by Dr Mohammad Hossein Amirhosseini, of the University of East London, Professor James Serpell of the University of Pennsylvania’s School of Veterinary Medicine, and the US dog training company Dogvatar who sponsored the research.

Dr Amirhosseini, of the School of Architecture, Computing and Engineering, said, 

This innovative AI-based methodology holds promise for revolutionising the selection and training of dogs for specific roles, both working and non-working.”

Nearly 50 per cent of people surrendering dogs to animal shelters in the USA cite behavioural problems as a contributory factor, with roughly a quarter citing such problems as the primary reason for giving up their pet.

While more research is needed to understand specific dog populations, by decoding the nuances of canine personalities, the research aims to foster stronger bonds between humans and their dogs, paving the way for more effective training programmes and tailored pet-owner interactions.

The report says, “These findings could provide a framework to explore personality matching between companion dogs and their owners and how this might contribute to the quality and durability of their relationships.

“The results of such studies could potentially generate insights regarding why dog-human partnerships succeed or fail, thereby reducing future rates of shelter relinquishment and euthanasia, and may also help to guide animal shelter and rescue groups towards more successful and mutually rewarding dog adoptions.”

Explore our BSc (Hons) Data Science and AI degree at UEL.