Inference Attack: Simple List Contents

This tool runs an attack based on simply requesting the database contents and checking for inference. This is probably the simplest inference attack, and works against pseudonymization, but not more sophisticated anonymization methods. Assuming that the table has N columns, this attack queries for N-1 columns and looks for cases where there is only a single value for the Nth column. With very high probability the N-1 columns identify a single user, and so by default there is only one value for the Nth column. We can then say that the Nth column is inferred by the values of the other N-1 columns. The tool is implemented in the python code, found in the repo

Leave a comment

Your email address will not be published.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


Other software tools

Tools for generating GDA Scores may be found on the GitHub repository

Utility Measure: Coverage

This tool measures the extent to which a given anonymization method hides data.

Find out more

Generate Defense Summary Graph

This tool generates a summary defense graph from a defense score.

Find out more

Utility Measure: Accuracy of Distinct User Counts

This tool measures the error introduced by the anonymization method for queries that count the number of distinct users.

Find out more