Use & Processing · Company practice
Machine learning and trained people detect and review violating content
Google says it uses a combination of people and machine-learning algorithms to detect policy-violating content and to help prevent others from finding it. The algorithms scan contributions for signals of suspicious user activity. Violating content is either removed by Google's automated models or flagged for further review by trained operators and analysts, who handle content evaluations that might be difficult for algorithms to perform alone.
Supporting quotations
We use a combination of people and machine-learning algorithms to detect such content and to help prevent others from finding it.
our machine-learning algorithms scan contributions for signals of suspicious user activity.
The policy-violating content is either removed by our automated models or flagged for further review by trained operators and analysts who conduct content evaluations that might be difficult for algorithms to perform alone.
Qualifications & exceptions
- Violating content is either removed automatically or flagged for human review; the policy does not say which cases go which way.
Not stated here
- The signals of suspicious user activity are not listed.
- The policy does not say whether removal by automated models can be reviewed by a person.