How to use the Data Anonymizer.
Removing obvious identifiers does not guarantee anonymity. Review combinations of fields and rare values that could still identify a person.
Make the workflow fit your task.
Identify direct identifiers, rare combinations and free text that may reveal identity. Agree a release purpose, then remove, generalize or replace fields accordingly. Keep any re-identification key separate and describe residual risk instead of labeling the output automatically anonymous.
- What you provide
- Small supplied dataset and chosen redaction policy.
- What you get
- Redacted or synthetic replacement values with residual-risk notes.
See the input and the result.
Illustrative input and output · a teaching example, not a live WebAct run
Example input
Fictional internal table: name Alex Example; email alex@example.com; role "sole night-shift laboratory coordinator". Prepare a public summary requiring only a broad job family.
Completed example
Public-summary draft: job_family: operations Removed: name, email and unique role wording. No stable person ID is needed for this purpose. A single-record summary could still be identifying through context; this transformation is not an anonymity certification.
Load this input into the prompt, then copy it to WebAct to try the task. Your result may differ from the illustration.
Decisions and troubleshooting.
Does replacing names with random IDs make a dataset anonymous?
It may only pseudonymize the records. Other fields or a retained mapping can still connect them to people.
Why can a person remain recognizable after removing email addresses?
Unique roles, dates, locations or narrative details can identify someone in combination. Review the whole record and intended audience.
Try it with your own source.
Replace the example with your material in the task prompt. Keep the requirements you need, then copy the task into WebAct.
Customize and copy the task ↑