The Incognito algorithm is the typical global algorithm, it obtain the best solution of minimum information loss in the whole situation by the distance vector of traversing the generalization. But the time efficiency of incognito algorithm is low, and it is not suitable for anonymizing amounts of data sets. In this paper, we propose a general framework which is aim to the global algorithm, at the same time, this paper introduce the property of law of identity to Incognito algorithm and put forward a new algorithm──F-Incognito. This algorithm regard one node which be with the highest generalization level as all nodes set, and judge the other attributes using the property that the result of the intersection of subset A and the universal set is equal to subset A, then this method could effectively reduces the calculation of amount of equivalence classes number. The experiment show that F-Incognito algorithm still have the characteristics of the global algorithm, at the same time, it also improved the generalized time efficiency and the time shorter ratio reach 60%. Therefore, F-Incognito algorithm has significant advantages when propose large data sets.
展开▼
