Japanese university releases AI prediction map on bear encounters

By Berk Kutay Gokmen

ISTANBUL (AA) - A research team from Sophia University in Tokyo has launched an online "Bear Encounter AI Prediction Map" designed to estimate the likelihood of human-bear interactions in 19 different regions, Kyodo News reported on Sunday.

The map uses AI to predict potential bear-related risks by analyzing local environmental data and reported sightings. The team aims for the tool to assist residents, tourists, and local authorities in preparing for bear activity, as bear attacks have become an increasing concern.

Led by Associate Professor Yusuke Fukazawa, a data science expert, the team created an AI-based tool that assesses the risk of bear-human encounters by examining sighting data from the past one to three years, as well as factors like forest conditions, road layouts, and population density.

Their analysis revealed higher risk areas along mountain foothills, river corridors where bears can move unnoticed, and remote valley roads, with aging communities facing an even greater risk of encounters.

On Oct. 20, the team launched a color-coded prediction map that divides each square kilometer into five levels of risk, ranging from low to very high, based on the estimated probability of bear presence, ranging from 0-20% to 80-100%.

The initiative was prompted by a record-high number of bear attack fatalities, 13 between April and Nov. 5, according to Japan's Environment Ministry.