Database Reconstruction Attacks (DRA)

1️⃣ version: 0.2.0

✍️ author: Mitchell Lisle

This notebook is adapted from a paper on database reconstruction attacks. You can find the paper here.

There are a number of reasons businesses and governments want to share information about people. It is important that when sharing information, you consider privacy and anonymity of the people that data is derived from. In most cases, aggregate data does little to hinder hackers from being able to re-create a database that is either very close, or exactly the same as the original data. In this application, we will take a simple example and re-create the database from nothing but aggregate statistics about those people.

We will mainly use Z3 for this.

Imagine we have the following database that contains information for people within a certain geographic area (going forward we refer to this area as a block.

We have 7 people in total in this block. Alongside age, we also have each resident's smoking status, employment status and whether they are married or not, we publish a variety of statistics. You have probably seen something similar in your countries census.

📓 To simplify the example, this fictional world has:

• Two marriage statuses; Married (True) or Single (False)
• Two smoking statuses; Non-Smoker (False) or Smoker (True)
• Two employment statuses; Unemployed (False) or Employed (True)

👾 One additional piece of logic we know is that any statistics with a count of less than 3 is suppressed

Install

You can run this application through a cli command. To install, you can run the following for inside the parent dir

pip install .

Usage

To run, you can use thte dra command after pip installing. The positional args are required:

1. input file: the file to pass in with the block stats
2. solution: the solution file which is what the original database. the input file stats are derived from this file
3. output file: the file to create for the re-constructed database to compare with the solution file

optional args:

• --min-age: the minimum age constraint, default: 0
• --max-age: the maximum age constraint, default: 115

dra data/block-stats.csv data/solution.csv data/reconstruction.csv

The logs will look something like this (Note the accuracy being reported when compared to the solution)

INFO:dra.logger:starting DatabaseReconstructionAttack runner...
INFO:dra.logger:running DRA...
INFO:dra.logger:DRA complete. Accuracy achieved 93%
INFO:dra.logger:reconstructed database: data/reconstruction.csv

The reconstructed database will look something like this:

age,married,smoker,employed
9,False,False,False
20,False,True,True
24,False,False,True
30,True,True,True
36,True,False,False
66,True,False,False
84,True,True,False

The original database:

age,married,smoker,employed
8,False,False,False
18,False,True,True
24,False,False,True
30,True,True,True
36,True,False,False
66,True,False,False
84,True,True,False