The western Japanese city of Otsu will begin next month the country's first attempt to use artificial intelligence to help teachers detect signs of serious school bullying.
AI will be used to analyze some 9,000 historical bullying cases reported by local elementary and junior high schools over the six years to this month. The system will also assess cases in which the bullying was only suspected.
The city signed an accord with Tokyo-based Hitachi Systems Ltd. on Friday to collaborate on the project.
The AI is expected to assist schools in detecting aggression in bullying cases, which can be triggered by a minor issue between students, even when some teachers lack knowledge or experience in identifying warning signs, according to the local education board.
Otsu Mayor Naomi Koshi said earlier she expects local schools to "act firmly against (bullying) without solely being dependent on teachers' experience, by having AI theoretically analyze past data."
Among the factors to be looked at via the AI are the grade-level of students impacted, their gender and the number involved, the timing of the bullying and where it occurred, as well as the students' academic records.
The education board believes the analysis, expected to be completed by October, will make known the characteristics of bullying to help teachers identify cases in their classrooms.
In Otsu, the suicide of a 13-year-old boy in 2011 was determined by a third-party panel two years later to have been caused by bullying.
Since then, the city's education board has required each school to report bullying cases within 24 hours.