Identifying the Risk of Attribute Disclosure by Mining Fuzzy Rules
In this paper we address the problem of controlling the disclosure of sensible information by inferring them by the other attributes made public. This threat to privacy is commonly known as prediction or attribute disclosure. Our approach is based on identifying those rules able to link sensitive information to the other attributes being released. In particular, the method presented in this paper is based on mining fuzzy rules. The fuzzy approach is compared to (crisp) decision trees in order to highlight pros and cons of it.
KeywordsSupport Vector Machine Fuzzy Rule Sensitive Information Privacy Protection Information Disclosure
Unable to display preview. Download preview PDF.
- 2.Joachims, T.: Making large-scale support vector machine learning practical. In: Smola, A., Scholkopf, B., Burges, C. (eds.) Advances in Kernel Methods: Support Vector Machines. MIT Press, Cambridge (1998)Google Scholar
- 3.Li, N., Li, T.: t-closeness: Privacy beyond k-anonymity and?-diversity. In: Proceedings of IEEE International Conference on Data Engineering (2007)Google Scholar
- 4.Machanavajjhala, A., Gehrke, J., Kifer, D., Venkitasubramaniam, M.: l-diversity: Privacy beyond k-anonymity. In: 22nd IEEE International Conference on Data Engineering (2006)Google Scholar
- 6.Pham, D.T., Bigot, S., Dimov, S.S.: Rules-f: a fuzzy inductive learning algorithm. In: Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, pp. 1433–1444 (2006)Google Scholar
- 7.Quinlan, R.: c4.5: Programs of machine learning (1993)Google Scholar
- 8.Ranilla, J., Luaces, O., Bahamonde, A.: A heuristic for learning decision trees and pruning them into classification rules. AICom (Artificial Intelligence Communication) 16(2) (2003) (in press)Google Scholar