id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_hiqnrmu3offw7blky2jcvqsw5i	Weijia Yang	A novel anonymization algorithm: Privacy protection and knowledge preservation	2010	12	.pdf	application/pdf	9463	1369	72	A novel anonymization algorithm: Privacy protection and knowledge preservation Unlike most anonymization methods, where data are generalized or permuted, our method anonymizes data by randomly breaking links among attribute values in records. Thus the data anonymized by our method maintains useful knowledge for statistical study. association between a value combination of Q-I and a value of sensitive data, then privacy can be compromised. I and the values of sensitive data, it is not necessary to anonymize methods, anonymizes a data set by randomly breaking the associations between its Q-I and sensitive data, we first define a statistical measurement of anonymity in Section 4.1. attribute) in a data set D and EntropyðQ iÞ be the value of the entropy For example, in the data set in Fig. 1, the two of its quasi-sensitive associations: When a data set is anonymized by the l-diversity algorithm, we the original data set D to get the quasi-sensitive associations AssoD,	./cache/work_hiqnrmu3offw7blky2jcvqsw5i.pdf	./txt/work_hiqnrmu3offw7blky2jcvqsw5i.txt