Granted Patent B2

US 11,755,584 · App. 16/033,814 · Granted Sep 12, 2023

Inventors: Christopher Poirel (Baltimore, MD); William Renner (Baltimore, MD); Eduardo Luiggi (Ellicott City, MD); Phillip Bracikowski (Indianapolis, IN)

Assignee: Forcepoint LLC

G06F16/24568G06F16/215G06F16/2365

Abstract

Claims (179)

1. A computer-implementable method for constructing a distribution of interrelated event features, comprising:

receiving a stream of data via a protected endpoint, the stream of data representing electronically-observable interactions by a user, the electronically-observable interactions being observed through at least one of an electronic device, a computer system and a software application executing on the computing system, the protected endpoint identifying a plurality of events from the interactions by the user, at least some of the plurality of events corresponding to user actions, the protected endpoint comprising an endpoint device and an endpoint agent, the endpoint agent executing on a hardware processor of the endpoint device, the protected endpoint providing a policy-based approach to network security;

extracting features from the plurality of events;

constructing a distribution of the features from the plurality of events, the distribution of the features being constructed via a scoring container update operation, the scoring container update operation using a scoring container to provide an approximation of a probability distribution of the features from the plurality of events for a particular time window across a sequence of time windows, the scoring container being implemented as a percentile container and a delta container, the percentile container collecting probability distribution of features extracted from certain interrelated events of a particular period of time which are then used to perform scoring operations, the delta container collecting probability distribution of features extracted from certain interrelated events of a particular period of time which are then used to update a persistent datastore;

performing a scoring operation using contents of the percentile container, the scoring operation being performed as:

P

ˆ

2

=

P

ˆ

1

+

∑

n

i

=

1

VAL

i

where P 1 corresponds to a first point in time;

P 2 corresponds to a second point in time; and,

VAL i corresponds to a score value of an associated feature;

analyzing the distribution of the features from the plurality of events;

generating a risk score for the user based upon the analyzing; and,

performing a risk assessment operation via a security analytics system based upon the distribution of features from the plurality of events, the security analytics system executing on a hardware processor, the security analytics system communicating with the endpoint via a network, the risk assessment operation taking into account the risk score, the protected endpoint communicating with the security analytics system via a network, the security analytics system being implemented in combination with the endpoint to perform operations associated with construction the distribution of interrelated event features.

2. The method of claim 1 , further comprising:

enriching data associated with each of the plurality of events prior to extracting features from the plurality of events.

3. The method of claim 2 , wherein:

the enriching data comprises

validating event data associated with at least some of the plurality of events,

disclaiming certain event data associated with at least some of the plurality of events;

deduplicating at least some of the plurality of events;

performing an entity resolution operation on at least some of the plurality of events;

performing an attachment enrichment operation on data associated with at least some of the plurality of events; and,

performing a domain enrichment on at least some of the plurality of events.

4. The method of claim 1 , further comprising:

labeling at least some of the plurality of events prior to extracting features from the plurality of events.

5. The method of claim 1 , wherein:

the extracting features comprises performing transformation operations on certain features associated with an event to generate a smaller set of derived features.

6. The method of claim 1 , further comprising:

processing a query relating to the plurality of events, the processing the query being performed via a streaming query framework.

7. The method of claim 1 , further comprising:

performing a container summation operation using contents of the delta container, the container summation operation being performed as:

Δ

=

∑

i

=

1

n

VAL

i

where VAL i corresponds to a score value of an associated feature.

8. The method of claim 7 , further comprising:

performing a scoring data update operation using the results of the scoring operation and the container summation operation, the scoring data update operation being performed as:

P 3 ={circumflex over (P)} 2 +Δ

where P 3 corresponds to a third point in time.

9. A system comprising:

a hardware processor;

a data bus coupled to the processor; and

a non-transitory, computer-readable storage medium embodying computer program code, the non-transitory, computer-readable storage medium being coupled to the data bus, the computer program code interacting with a plurality of computer operations and comprising instructions executable by the processor and configured for:

receiving a stream of data via a protected endpoint, the stream of data representing electronically-observable interactions by a user, the electronically-observable interactions being observed through at least one of an electronic device, a computer system and a software application executing on the computing system, the protected endpoint identifying a plurality of events from the interactions by the user, at least some of the plurality of events corresponding to user actions, the protected endpoint comprising an endpoint device and an endpoint agent, the endpoint agent executing on a hardware processor of the endpoint device, the protected endpoint providing a policy-based approach to network security;

extracting features from the plurality of events;

constructing a distribution of the features from the plurality of events, the distribution of the features being constructed via a scoring container update operation, the scoring container update operation using a scoring container to provide an approximation of a probability distribution of the features from the plurality of events for a particular time window across a sequence of time windows, the scoring container being implemented as a percentile container and a delta container, the percentile container collecting probability distribution of features extracted from certain interrelated events of a particular period of time which are then used to perform scoring operations, the delta container collecting probability distribution of features extracted from certain interrelated events of a particular period of time which are then used to update a persistent datastore;

performing a scoring operation using contents of the percentile container, the scoring operation being performed as:

P

ˆ

2

=

P

ˆ

1

+

∑

n

i

=

1

VAL

i

where P 1 corresponds to a first point in time;

P 2 corresponds to a second point in time; and,

VAL i corresponds to a score value of an associated feature;

analyzing the distribution of the features from the plurality of events;

generating a risk score for the user based upon the analyzing; and,

performing a risk assessment operation via a security analytics system based upon the distribution of features from the plurality of events, the security analytics system executing on the hardware processor, the security analytics system communicating with the endpoint via a network, the risk assessment operation taking into account the risk score, the protected endpoint communicating with the security analytics system via a network, the security analytics system being implemented in combination with the endpoint to perform operations associated with construction the distribution of interrelated event features.

10. The system of claim 9 , wherein the instructions are further configured for:

enriching data associated with each of the plurality of events prior to extracting features from the plurality of events.

11. The system of claim 10 , wherein:

the enriching data comprises

validating event data associated with at least some of the plurality of events,

disclaiming certain event data associated with at least some of the plurality of events;

deduplicating at least some of the plurality of events;

performing an entity resolution operation on at least some of the plurality of events;

performing an attachment enrichment operation on data associated with at least some of the plurality of events; and,

performing a domain enrichment on at least some of the plurality of events.

12. The system of claim 9 , wherein the instructions are further configured for:

labeling at least some of the plurality of events prior to extracting features from the plurality of events.

13. The system of claim 9 , wherein:

the extracting features comprises performing transformation operations on certain features associated with an event to generate a smaller set of derived features.

14. The system of claim 9 , wherein:

processing a query relating to the plurality of events, the processing the query being performed via a streaming query framework.

15. The system of claim 9 , wherein the instructions are further configured for:

performing a container summation operation using contents of the delta container, the container summation operation being performed as:

Δ

=

∑

i

=

1

n

VAL

i

where VAL i corresponds to a score value of an associated feature.

16. The system of claim 15 , wherein the instructions are further configured for:

performing a scoring data update operation using the results of the scoring operation and the container summation operation, the scoring data update operation being performed as:

P 3 ={circumflex over (P)} 2 +Δ

where P 3 corresponds to a third point in time.

17. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:

receiving a stream of data via a protected endpoint, the stream of data representing electronically-observable interactions by a user, the electronically-observable interactions being observed through at least one of an electronic device, a computer system and a software application executing on the computing system, the protected endpoint identifying a plurality of events from the interactions by the user, at least some of the plurality of events corresponding to user actions, the protected endpoint comprising an endpoint device and an endpoint agent, the endpoint agent executing on a hardware processor of the endpoint device, the protected endpoint providing a policy-based approach to network security;

extracting features from the plurality of events;

constructing a distribution of the features from the plurality of events, the distribution of the features being constructed via a scoring container update operation, the scoring container update operation using a scoring container to provide an approximation of a probability distribution of the features from the plurality of events for a particular time window across a sequence of time windows, the scoring container being implemented as a percentile container and a delta container, the percentile container collecting probability distribution of features extracted from certain interrelated events of a particular period of time which are then used to perform scoring operations, the delta container collecting the probability distribution of features extracted from the certain interrelated events of the particular period of time which are then used to update a persistent datastore;

performing a scoring operation using contents of the percentile container, the scoring operation being performed as:

P

ˆ

2

=

P

ˆ

1

+

∑

n

i

=

1

VAL

i

where P 1 corresponds to a first point in time;

P 2 corresponds to a second point in time; and,

VAL i corresponds to a score value of an associated feature;

analyzing the distribution of the features from the plurality of events;

generating a risk score for the user based upon the analyzing; and,

performing a risk assessment operation via a security analytics system based upon the distribution of features from the plurality of events, the security analytics system executing on a hardware processor, the security analytics system communicating with the endpoint via a network, the risk assessment operation taking into account the risk score, the protected endpoint communicating with the security analytics system via a network, the security analytics system being implemented in combination with the endpoint to perform operations associated with construction the distribution of interrelated event features.

18. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are further configured for:

enriching data associated with each of the plurality of events prior to extracting features from the plurality of events.

19. The non-transitory, computer-readable storage medium of claim 14 , wherein:

the enriching data comprises

validating event data associated with at least some of the plurality of events,

disclaiming certain event data associated with at least some of the plurality of events;

deduplicating at least some of the plurality of events;

performing an entity resolution operation on at least some of the plurality of events;

performing an attachment enrichment operation on data associated with at least some of the plurality of events; and,

performing a domain enrichment on at least some of the plurality of events.

20. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are further configured for:

labeling at least some of the plurality of events prior to extracting features from the plurality of events.

21. The non-transitory, computer-readable storage medium of claim 13 , wherein:

the extracting features comprises performing transformation operations on certain features associated with an event to generate a smaller set of derived features.

22. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are further configured for:

processing a query relating to the plurality of events, the processing the query being performed via a streaming query framework.

23. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are deployable to a client system from a server system at a remote location.

24. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are provided by a service provider to a user on an on-demand basis.

25. The non-transitory, computer-readable storage medium of claim 13 , wherein the computer executable instructions are further configured for:

performing a container summation operation using contents of the delta container, the container summation operation being performed as:

Δ

=

∑

i

=

1

n

VAL

i

where VAL i corresponds to a score value of an associated feature.

26. The non-transitory, computer-readable storage medium of claim 25 , wherein the computer executable instructions are further configured for:

performing a scoring data update operation using the results of the scoring operation and the container summation operation, the scoring data update operation being performed as:

P 3 ={circumflex over (P)} 2 +Δ

where P 3 corresponds to a third point in time.

Assignments (9)