Metrics for Anonymity

This paper proposes the use of a system of measures for Anonymity based on the characteristics that define their main properties: the index of uniformity of the probability distribution and size of all anonymous. In previous proposals, the most widely used is based on the entropy, an index used in information theory, which has several drawbacks with respect to representation of the properties mentioned in the first place does not represent them explicitly and being a logarithmic index does not represent Anonymity levels or degrees that require linear behavior. To measure the uniformity index is proposed to use the Root Mean Square Error criterion or the Jensen-Shannon divergence. For anonymous set size proposes the use of a function of N.