Privacy and Security in Data Analysis (CS685-003)
NOTE TO STUDENTS: This class (CS685-003)
"Privacy and Security in Data Publishing and Data Analysis" is
It will deal with some computational techniques commonly used
for privacy preserving and security enhancing in
data publishing and data analysis. It will
cover some very new techniques and developments in this emerging
area, such as privacy preserving in social network analysis
and dual privacy protection. Knowledge in data mining is helpful,
but not necessary.
Semester: Spring, 2017.
Class Time: MWF: 11:00AM-11:50AM.
Classroom: FP Anderson Tower 267.
Instructor: Jun Zhang, E-mail:firstname.lastname@example.org, Tel:257-3892.
Office: 321 Marksbury building.
Office Hours: MW: 9:00am - 10:00am, and by appointment.
Students must have taken at least two 500 level CS courses
at the University of Kentucky before eligible for registration
for this course. Students do not meet this prerequisite should
meet with the instructor before a registration can be granted.
Suggested Text Book:
No Formal Text Book
There is an electronic book in a zipped file, and
a list of PPDM reference papers.
Data collection and data analysis have become ubiquitous
in modern world. Along with this trend, the need to
protect private and sensitive information in data has
become an important issue.
This course will study a few state-of-the-art techniques
in protecting data privacy and data security when the data
is released to public or is subject to computer-based
analysis, such as data mining. The contents include some of the
Ph.D. thesis research results of the previous students.
1.) Brief introduction to data mining;
2.) Privacy-preserving data mining;
3.) Tabular privacy-preserving publishing;
4.) Matrix decomposition in privacy-preserving data mining;
5.) Wavelet analysis in privacy-preserving data analysis;
6.) Privacy attacks;
7.) Privacy-preserving in social network analysis.
8.) Introduction to on-line recommendation systems.