17 March 2021

Research pick: Predicting canine pack patterns - "Map simulation of dogs’ behaviour using population density of probabilistic model"

Human understanding of animal behaviour is important not only from a purely scientific perspective but also from the perspective of disease prevention and control. This is especially poignant when considering those animals of vectors of disease that can be transmitted to humans and perhaps even underpin the emergence of novel pathogens such as the SARS-CoV-2 virus which has led to the current global Covid-19 pandemic.

Writing in the International Journal of Computer Applications in Technology, a team from Thailand has looked at canine behaviour and how improved understanding might help in rabies control through better animal vaccination programs. Better understanding might also be useful in understanding behaviour when there is a major outbreak. The team has modelled the behaviour of individual dogs and packs (canine communities) and the way in which individual animals may explore new territory. They look closely at the “tie-strength” between any two dogs. They have validated their model on a region of the island of Saibai in northwestern Torres Strait islands, Australia. Saibai lies off the south-eastern coast of New Guinea.

The simulated data fit with actual tracking data to within just over 6 percent accuracy in terms of tie-strength. As such, the team has now simulated canine behaviour in three Thai cities and demonstrated a difference in how tie-strength affects behaviour. This, they suggest, may reflect significantly higher average numbers of dogs in a given area, the larger group distances and bigger connections between dogs and their packs.

The team suggests that the same approach to modelling canine behaviour might be extended to the walking behaviour of other animals with relative ease.

Jiwattanakul, J., Youngjitikornkun, C., Kusakunniran, W., Wiratsudakul, A., Thanapongtharm, W. and Leelahapongsathon, K. (2021) ‘Map simulation of dogs’ behaviour using population density of probabilistic model’, Int. J. Computer Applications in Technology, Vol. 65, No. 1, pp.14–24.

No comments: