DPP Topic 3 - Data Anonymization

Question 1

Data Anonymization

Answer 1

The process of removing or modifying personally identifiable information from a dataset to prevent the identification of individuals.

Question 2

Original Database

Answer 2

The initial dataset containing personal information.

Question 3

Published Database

Answer 3

The anonymized dataset that is made available for use.

Question 4

Anonymized Data

Answer 4

The modified data that retains the usefulness of the original data while protecting individual privacy.

Question 5

Original Data

Answer 5

The personal information in the initial dataset.

Question 6

Balancing Data Privacy and Data Utility

Answer 6

The goal of anonymization is to make data less specific while retaining its usefulness.

Question 7

1. Determine release model
   (Data Preparation)

Answer 7

Deciding whether the anonymized dataset will be made public or kept non-public.

Question 8

2. Determine re-identification risk threshold (Data Preparation)

Answer 8

Higher risk thresholds lead to increased data anonymity but decreased data utility.

Question 9

3. Classify data attributes
   (Data Preparation)

Answer 9

Identifying explicit identifiers, quasi-identifiers, and sensitive data.

Question 10

4. Remove unused data attributes
   (Data Preparation)

Answer 10

Suppressing attributes that are not required in the anonymized dataset.

Question 11

5. Anonymize identifiers
   (Data Execution)

Answer 11

Applying relevant anonymization techniques to different types of identifiers.

Question 12

6. Evaluate the solution
   (Data Execution)

Answer 12

Assessing the anonymized dataset for sufficient data anonymity and utility.

Question 13

7. Determine controls required
   (Data Execution)

Answer 13

Implementing technical and non-technical controls to protect the anonymized data.

Question 14

8. Document anonymization process
   (Data Execution)

Answer 14

Recording the details of the anonymization process for future reference.

Question 15

1) Attribute Suppression
(Techniques)

Answer 15

Removal of an entire attribute (column) from the dataset. (Strongest Technique)

Question 16

2) Character Masking
(Techniques)

Answer 16

Replacing characters in a data value with a symbol (e.g., “*” or “x”).

Question 17

3) Generalization
(Techniques)

Answer 17

Reducing the precision of data by converting it into a range of values.

Question 18

4) Swapping
(Techniques)

Answer 18

Rearranging data values within the dataset to break the link between individual records.

Question 19

5) Data Perturbation
(Techniques)

Answer 19

Modifying the original data values to be slightly different.

Question 20

6) Synthetic Data
(Techniques)

Answer 20

Artificially creating data that captures the statistical distributions of the original data.

Question 21

7) Data Aggregation
(Techniques)

Answer 21

Converting a dataset from individual records to summarized values.

Question 22

8) K-anonymity
(Techniques)

Answer 22

A property of a dataset where each combination of values appears in at least k records, making it harder to identify individuals.

Question 23

9) Pseudonymization

Answer 23

Replacing identifying data with made-up values that have no relationship to the original values.