MSc Data Science (Social Analytics) / Course details

The course unit aims to introduce basic and advanced concepts of privacy, confidentiality and statistical disclosure control, and to provide an understanding of the evolution of research and developments in these areas. The topics covered include: (1) introduction to statistical disclosure control, disclosure risk scenarios and types of disclosure risks; (2) the anonymisation decision making framework; (3) measuring disclosure risk for different formats of outputs; (3) common statistical disclosure control methods and their impact on data quality and utility; (4) differential privacy and other mathematically rigorous forms of formal privacy; (5) synthetic data. 

The lectures will be spread out over 10 weeks, with 3 hours per week and will include a 1.5-hour lecture and a 1.5-hour practical (or 2-hour lecture and 1 hour practical depending on the topic).

The topics covered in each week will be:  

Introduction; Understanding Privacy and Confidentiality

Anonymisation 1

Anonymisation 2

Statistical disclosure control (SDC), measuring disclosure risk, data utility, R-U confidentiality map  

SDC Methods 1

SDC Methods 2

K Anonymity, t-closeness, l-diversity, Differential Privacy 1

Differential Privacy 2

Data Synthesis 1

Data Synthesis 2; Review 

The lectures will be spread out over 10 weeks, with 3 hours per week and will include a 1.5 hour lecture and a 1.5 hour practical (or 2 hour lecture and 1 hour practical depending on the topic). 

Understand the anonymisation decision making framework

Understand Statistical Disclosure Control including how to apply, evaluate and critique different methods depending on the statistical output with respect to disclosure risk and data utility

Understand formal privacy methods, including differential privacy, and their applications

Generate synthetic data and understand their potential and limitations 

Process datasets in preparation for privacy assessments

Understand the differences between statistical disclosure control methods and formal privacy and critically evaluate the different approaches with respect to their commonalities and differences

Construct and appraise disclosure risk and data utility measures 

Write reports for non-academic audiences applying the anonymisation framework to statistical outputs

Present on an application of statistical disclosure control 

Demonstrate computing skills to plan and implement small projects applying statistical disclosure control and formal privacy to different types of outputs

Present material on privacy for a wider audience