Sunday, April 18, 2010

My Research Paper - Comparison between 2 models of K-Anonymity: Incognito and Mondrian

TOPIC

Comparison between 2 models of K-Anonymity: Incognito and Mondrian

ABSTRACT

Often organizations publish microdata for purposes such as public health and demographic research. Although attributes that could identify the individuals such as names and IDs are removed, by combining information (such as Zipcode and Birthdays) within different database, these individuals could still be identified through “Joining Attacks” – Combining 2 different database by cross-linking the similar information.

K-Anonymity has been proposed as a mechanism for protecting privacy by generalizing or suppressing certain portion of the released microdata. There had been several models of K-Anonymity since, each release attempts to improve the previous version’s weaknesses. In this paper, we will compare 2 models of K-Anonymity, namely Incognito (full-domain generalization) and Mondrian (multidimensional model), in terms of their effectiveness as well as their performance.

GENERAL TERMS

Algorithms, Experimentation, Theory

KEYWORDS

K-Anonymity, Incognito, Mondrian

REFERENCES

[1] Kristen LeFevre, David J. DeWitt, Raghu Ramakrishnan, 2005. Incognito: Efficient FullDomain KAnonymity. University of Wisconsin, Madison. http://www.cse.iitb.ac.in/dbms/Data/Courses/CS632/Papers/incognito.pdf
[2] Kristen LeFevre, David J. DeWitt, Raghu Ramakrishnan, 2006. Mondrian Multidimensional K-Anonymity. University of Wisconsin, Madison.
http://www.cse.iitb.ac.in/dbms/Data/Courses/CS632/Papers/kanon-lefevre-icde06.pdf
[3] C. Blake and C. Merz. UCI repository of machine learning databases, 1998.
[4] Wikipedia Authors, Visual Basic for Applications, Wikipedia, Retrieved on 6 April 2010.
http://en.wikipedia.org/wiki/Visual_Basic_for_Applications
[5] Wolfram Research Inc, Box-Muller Transformation, Wolfram MathWorld, Retrieved on 6 April 2010.
http://mathworld.wolfram.com/Box-MullerTransformation.html
[6] R. Bayardo and R. Agrawal. Data privacy through optimal k-anonymization. In ICDE, 2005.
[7] L. Sweeney. K-Anonimity: A model for protecting privacy, International Journal on Uncertainty, Fuzziness, and Knowledge-based Systems, 10(5):557-570, 2002

No comments:

Post a Comment

Do provide your constructive comment. I appreciate that.