System and method for intelligent agents for decision support in network identity graph based identity management artificial intelligence systems
Systems and methods for embodiments of a graph based artificial intelligence systems for identity management are disclosed. Embodiments of the identity management systems disclosed herein may utilize a network graph approach to analyzing identities or entitlements of a distributed networked enterprise computing environment. Specifically, in certain embodiments, an artificial intelligence based identity governance systems may include an intelligent decision support agent to provide an approval or denial recommendation for an access request. To provide an approval or denial recommendation, the intelligent agent may utilize a classifier trained on historical certification data. The intelligent agent may utilize features which represent relevant signals to the approval or denial decision including features that may be associated with a network graph of the identities and entitlements of the enterprise computing environment.
1. A non-transitory computer readable medium, comprising instructions for:
obtaining historical certification data for an enterprise, the historical certification data comprising a set of historical certification decisions, each historical certification decision comprising a historical access request including an identity and an entitlement, an approval or denial decision for the historical access request, and a value for a feature determined from an identity graph associated with the enterprise;
training a classifier with the historical certification data;
receiving an access request including an identity and an entitlement and a request for a decision for the access request;
enhancing the received access request with a value for the feature based on the identity or entitlement of the access request and the identity graph associated with the enterprise;
submitting the enhanced access request to the classifier;
receiving a decision for the enhanced access request from the classifier; and
returning the decision for the access request to a user of the identity management system.
2. The non-transitory computer readable medium of claim 1 , wherein the identity graph associated with the enterprise is generated by:
creating a node of the identity graph for each of a set of identities determined from identity management data of the enterprise, the identity management data comprising data on the set of identities and a set of entitlements associated with the set of identities utilized in identity management in the enterprise;
for each first identity and second identity that share at least one entitlement of the set of entitlements, creating an edge of the identity graph between a first node representing the first identity and a second node of the identity graph representing the second identity, and
generating a similarity weight for each edge of the identity graph between each first node and second node based on a number of the set of entitlements shared between the first identity represented by the first node and the second identity represented by the second node.
3. The non-transitory computer readable medium of claim 1 , wherein the feature is a primary feature or a composite feature.
4. The non-transitory computer readable medium of claim 1 , wherein the identity graph used to determine the value for the feature of each historical certification decision is a version of the identity graph closest in time to the historical access request of that historical certification decision.
5. The non-transitory computer readable medium of claim 1 , wherein the historical certification data was determined using a rules based classification agent.
6. The non-transitory computer readable medium of claim 5 , wherein the rules based classification agent is utilized until the classifier meets a performance metric.
7. The non-transitory computer readable medium of claim 1 , wherein the instructions further comprise instructions for:
receiving a request for an interpretation of the decision for the access request;
determining a top set of features that resulted in the decision based on the classifier; and
returning the top set of features to the user.
8. The non-transitory computer readable medium of claim 7 , wherein determining the top set of features comprise querying an interpreter to determine the top set of features from a local model.
9. The non-transitory computer readable medium of claim 8 , wherein the local model was built by the interpreter by querying the classifier in a neighborhood of the access request.
10. The non-transitory computer readable medium of claim 9 , wherein the interpreter is independent from the training of the classifier.
11. An identity management system, comprising:
a processor; and
a non-transitory, computer-readable storage medium, including computer instructions for:
training a model with the historical certification data for an enterprise, the historical certification data comprising a set of historical certification decisions, each historical certification decision comprising a historical access request including an identity and an entitlement, an approval or denial decision for the historical access request, and a value for a feature determined from an identity management data associated with the enterprise;
receiving an access request including an identity and an entitlement and a request for a decision for the access request;
determining a value for the feature for the received access request based on the identity or entitlement of the access request and the identity management data associated with the enterprise;
submitting the access request and the determined value to the model;
receiving a decision for the access request from the model; and
returning the decision for the access request to a user of the identity management system.
12. The system of claim 11 , wherein the identity management data includes an identity graph associated with the enterprise.
13. The system of claim 12 , wherein the identity graph includes nodes representing identities and entitlements.
14. The system of claim 11 , wherein the identity graph used to determine the value for the feature of each historical certification decision is a portion of the identity graph representing a state closest in time to the historical access request of that historical certification decision.
15. The system of claim 11 , wherein the model is a classifier.
16. An method for identity management, comprising:
training a model with the historical certification data for an enterprise, the historical certification data comprising a set of historical certification decisions, each historical certification decision comprising a historical access request including an identity and an entitlement, an approval or denial decision for the historical access request, and a value for a feature determined from an identity management data associated with the enterprise;
receiving an access request including an identity and an entitlement and a request for a decision for the access request;
determining a value for the feature for the received access request based on the identity or entitlement of the access request and the identity management data associated with the enterprise;
submitting the access request and the determined value to the model;
receiving a decision for the access request from the model; and
returning the decision for the access request to a user of the identity management system.
17. The method of claim 16 , wherein the identity management data includes an identity graph associated with the enterprise.
18. The method of claim 17 , wherein the identity graph includes nodes representing identities and entitlements.
19. The method of claim 16 , wherein the identity graph used to determine the value for the feature of each historical certification decision is a portion of the identity graph representing a state closest in time to the historical access request of that historical certification decision.
20. The method of claim 16 , wherein the model is a classifier.