Detecting and censoring sensitive data
Under Dutch law (refer to “WOB”), civilians have the right to request information about all proceedings of the government. When a request is granted, the relevant documents have to
be disclosed. Therefore, it is important that these documents do not contain any privacy-sensitive data.
Gemeente Utrecht approached us to see (1) if the sensitive information in documents can be detected and (2) if those documents can be censored automatically.
We found that most sensitive information can be detected automatically. However, detecting 100% of the sensitive information is infeasible. And, except for PDF files and images, most document types can be censored automatically.
Want to know more about the project? Feel free to contact us.
|Detecting and censoring sensitive data.pdf|