A Japanese device predicts employee resignations

Employers worried about employee turnover or wondering how long they'll be on the job soon can believe this. Artificial intelligence (AI) to identify which employee is likely to leave the organization next.

This tool was developed by Japanese researchers to help companies provide targeted support to their employees, They should be prevented from resigning.

Employee data is analyzed using it, from commuting to work to personal information such as age and gender The tool was developed by a professor at Tokyo City University With a new local company.

The tool also analyzes data from former employees who have left the company.

From all this data, the computer predicts the resignation rate of new hires as a percentage, explains Professor Naruhiko Shiratori to AFP.

“We're in Test phase “With this AI tool across multiple companies, they create a model for each,” says the researcher.

Employers can use the results to determine whether the company is willing to support a “high-risk (resigning) employee — without showing them raw results that might surprise them.” Because the AI ​​predicts that “you will face difficulties”.” says Shiratori.

To develop this tool, the researchers built on a previous study that used AI to predict the profile of college students.

Japanese companies traditionally hire young graduates all at once in April each year.

But about 10% of these new workers They abandon their posts in the first year, and about 30% after three years, according to government data.

As population declines in the Japanese archipelago, Japanese companies are increasingly seeking to take care of their young employees. creates a labor shortage In many fields of activity.

See also  Windows 10X has been released on the Internet. Take a look at what innovative features it combines and what limitations it comes with

Misty Tate

"Freelance twitter advocate. Hardcore food nerd. Avid writer. Infuriatingly humble problem solver."

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top