| CPC H04L 63/20 (2013.01) [H04L 41/28 (2013.01); H04L 63/101 (2013.01)] | 20 Claims |
|
1. A method for facilitating a permission management pipeline to provide policy taxonomy governance, the method being implemented by at least one processor, the method comprising:
receiving at least one data set that is shared from a producer application, the at least one data set including data, corresponding metadata, and producer policy logic;
training a machine learning (ML) model based on a cross-validation technique, a holdout technique, and a bootstrap technique, the ML model being trained in pattern detection correlating with a policy taxonomy governance associated with different computing platforms and the ML model being trained in operating with a least squares error rate within a predetermined range;
tagging the data and the corresponding metadata for each of the at least one data set to identify raw data and permission schema data for input into the trained ML model;
determining, by the trained ML model, at least one rule by parsing the producer policy logic, the at least one rule including a declarative rule;
generating, by the trained ML model, at least one standard policy based on the determined at least one rule, the at least one standard policy corresponding to an executable standard entitlement policy;
executing the generated at least one standard policy by using a policy enforcement function, the tagged data, and the tagged corresponding metadata; and
generating based on a result of the executing, at least one producer permission that is paired with the at least one data set, the at least one producer permission including an entity-group access-control list of permissions that is associated with the at least one data set.
|