Partitions and pignistic distributions in GIT

The Ambiguity measure is based on a probability distribution derived from a body of evidence using pignistic techniques. The Aggregate Uncertainty measure also creates a probability distribution from a body of evidence but this distribution is chosen to maximize the Shannon entropy. A body of evidence can be used to form a partition of the underlying domain. When we can group the terms of the Ambiguity or Aggregate Uncertainty using this partition, some interesting connections between entropies and nonspecificities are revealed.