Apparatus and Methods For Anonymizing a Data Set

摘要

Methods and systems are disclosed for anonymizing a dataset that correlates a set of entities with respective attributes. The method comprises determine clusters of similar entities. Determining the clusters comprises (1) partitioning the entities into a first group with similar attributes to one another and a complement group of entities with similar attributes to one another and (2) recursively repeating the partitioning on the groups until every group meets one or more criteria. The partitioning a group comprises choosing a reference entity from the group, determining a symmetric set of attributes based on the reference entity attributes and on an average of the group's attributes, and assigning each entity to the first or second group depending on whether it's attributes are more similar to those of the reference user or to those of the symmetric set.