id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-025519-265qdtw6	Zouinina, Sarah	A Two-Levels Data Anonymization Approach	2020-05-06		.txt	text/plain	3486	222	52	Consequently, privacy preservation through machine learning algorithms were designed based on cryptography, statistics, databases modeling and data mining. To this purpose, we revisited all the previously proposed approaches, and we added a second level of anonymization by incorporating the discriminative information and using Adaptive Weighting of Features to improve the quality of the anonymized data. The paper is organised into four sections: the first dresses the different approaches of privacy preserving using machine learning, the second sums up the previously proposed approaches, the third discusses the introduction of the discriminative information and the fourth validates the method experimentally on six different datasets. The two models propose an algorithm that relies on the classical Self Organizing Maps (SOMs) [10] and collaborative Multiview clustering in purpose to provide useful anonymous datasets [9] . As shown in the Table 5 , the introduction of the discriminant information improves the utility of the anonymized datasets for all of the methods proposed.	./cache/cord-025519-265qdtw6.txt	./txt/cord-025519-265qdtw6.txt