Maintaining the Quality and Reliability of Employee Review Data
First written in FY2022. 3 of 4 entries say "Restated". Latest entry: 2026-03-24.
No other company in this theme has enough history to compare yet.
- 2026-03-24RestatedManagement Planning
「“500 characters or more” report entry rule and response guidelines / visual review of all reviews by operating staff / approximately 7% of posts detected and unpublished as fraudulent posts, including promotional or prank posts」
The quality controls involving review submission standards, AI and visual screening, and the unpublishing of fraudulent posts were reiterated. - FY2025RestatedAnnual report
「For every employee review submitted, we conduct AI-based automated screening and visual review by dedicated staff. We have established operating procedures and a review framework to prevent posts that violate laws or constitute defamation from being published on the service, and work to improve not only the quantity but also the quality of published employee reviews and other data.」
Data-quality controls, including AI-based automated screening and visual review by dedicated staff, remain in operation. - FY2024RestatedAnnual report
「For every employee review submitted, we conduct AI-based automated screening and visual review by dedicated staff. We have established operating procedures and a review framework to prevent posts that violate laws or constitute defamation from being published on the service, and work to improve not only the quantity but also the quality of published employee reviews and other data.」
Data-quality controls, including AI-based automated screening and visual review by dedicated staff, remain in operation. - FY2022First statedAnnual report
「For every employee review submitted, we conduct AI-based automated screening and visual review by dedicated staff. We have established operating procedures and a review framework to prevent posts that violate laws or constitute defamation from being published on the service, and work to improve not only the quantity but also the quality of published employee reviews and other data.」
The company described measures to improve data integrity and reliability, including a minimum length of 500 characters, employment-period requirements, AI screening and visual review.