Permission level generation based on adaptive learning

摘要

This disclusre teaches a clustering system. The clustering system includes an initializr, a mergerer, a clustering asignment selector and an assigner database. The initializer adapted to perform at least one of: update of self-similarity values and calculation of an initial score for an optional clustering assignment. The mergerer is adapted to perform at least one of: calculate the delta values for the potential merger of any of two clusters, and update the pairwise similarity as a result of merging of two clusters with the highest score in each round of calculations. The clustering assignment selector is adapted to determine the clustering assignment with the highest overall score. The assigner database is adapted to store and retrieve at least: pairwise similarity and updates thereof, and scores of potential clustering assignment determined at each round of calculations.