Systems and methods for intelligently implementing a machine learning-based digital threat mitigation service

• US 10,623,423 B1
• Filed: 06/06/2019
• Issued: 04/14/2020
• Est. Priority Date: 06/06/2019
• Status: Active Grant

• Current Assignee: Sift Science, Inc.
• Original Assignee: Sift Science, Inc.
• Inventors: Sadaghiani, Fred, Mann, Megan, Lopatin, Aleksandr, Grant, Noah
• Primary Examiner(s): Edwards, Linglan E
• Application Number:
  US16/433,934
• Time in Patent Office: 313 Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 20/20 Ensemble learning
• CPC Class Codes: H04L 63/1416 Event detection, e.g. attack signature detection
• CPC Class Codes: H04L 63/1425 Traffic logging, e.g. anomaly detection
• CPC Class Codes: H04L 63/1433 Vulnerability analysis
• CPC Class Codes: H04L 63/1441 Countermeasures against malicious traffic countermeasures against attacks on cryptographic mechanisms H04L9/002
• CPC Class Codes: H04L 63/1466 Active attacks involving interception, injection, modification, spoofing of data unit addresses, e.g. hijacking, packet injection or TCP sequence number attacks

Abstract

Systems and methods include implementing a review queue interface that includes: a review queue comprising a listing of distinct review items; a current state for each of distinct review items; a listing for each review item of the distinct review items of one or more client browsers that are interacting with each review item; identifying client browser activity of the one or more client browsers; computing a computed state for each of distinct review items based on the client browser activity; computing changes to the state of review items based on an assessment of the current state and the computed state for each of distinct review items; and automatically updating a state of one or more of the distinct review items within the review queue interface based on a difference between the current state and the computed state of the one or more of the distinct review items.

Claims

1. A machine learning-based system implementing a collisionless queue, the system comprising:

a distributed network of computers, a machine learning-based service being implemented by the distributed network of computers to;

implement a review queue interface that includes a display of;

(i) a review queue comprising a listing of a plurality of distinct review items;

(ii) a current state for each of the plurality of distinct review items, wherein the current state is arranged proximate to each of the plurality of distinct review items; and

(iii) for each review item of the plurality of distinct review items, a listing of distinct browser identifiers for each of one or more client browsers that are interacting with each respective review item;

records to a persistent log of events client browser activity data of the one or more client browsers interacting with the plurality of distinct review items;

computes a computed state for each of the plurality of distinct review items based at least on the client browser activity data, wherein computing the computed state includes;

replaying the persistent log of events;

identifying one or more instances in which the one or more client browsers is viewing one or more of the plurality of distinct review items based on the replay of the persistent log of events;

identifying one or more instances in which the one or more client browsers has viewed away from the one or more of the plurality of distinct review items based on the replay of the persistent log of events;

computes changes to the current state that is displayed for each of the plurality of distinct review items based on an assessment of the current state and the computed state for each of the plurality of distinct review items; and

automatically updates a state of one or more of the plurality of distinct review items within the review queue interface based on identifying a difference between the current state and the computed state of the one or more of the plurality of distinct review items.

2. The system according to claim 1, wherein

a review item relates to an item comprising one or more of digital activity data and digital transaction data sourced from an automated decisioning workflow of the machine learning-based service without a disposal decision.

3. The system according to claim 1, wherein

the machine learning-based service automatically pushes a notification to any one of the one or more client browsers interacting with a review item with the state change based on the difference between the current state and the computed state of the one or more of the plurality of distinct review items.

4. The system according to claim 1, wherein

the current state or the computed state for a given review item of the plurality of review items includes one or more of:

[1] unclaimed state;

[2] claimed state;

[3] decided state; and

[4] expired state.

5. The system according to claim 4, wherein

a review item having a claimed state relates to a review item in a review queue that is being viewed by or via at least one client browser according to one or more viewing semantics.

6. The system according to claim 4, wherein

a review item having an unclaimed state relates to a review item that is not being viewed by or via at least one browser.

7. The system according to claim 4, wherein

a review item having a decided state relates to a review item having an associated disposal decision or relates to a review item that has been decided on by an analyst associated with a browser.

8. The system according to claim 4, wherein

a review item having an expired state relates to a review item without a disposition or has not been explicitly decided on and that has passed a maximum amount of time in which the item may be made or shown as available via the review queue interface.

9. The system according to claim 1, wherein:

the machine learning-based service assesses a predetermined expiry associated with each of the plurality of distinct review items;

computes changes to the state for one or more of the plurality of review items based on the assessment of the predetermined expiry; and

automatically updating the state of one or more of the plurality of distinct review items within the review queue interface based on an expiration of the predetermined expiry of the one or more of the plurality of distinct review items.

10. The system according to claim 1, wherein:

the machine learning service;

records within a persistent log of events client browser activity of the one or more client browsers interacting with the review queue interface; and

computing the computed state for each of the plurality of distinct review items is based on a replay of the persistent log of events.

11. The system according to claim 1, wherein

the computed state of a review item of the plurality of distinct review items is updated based on a client browser of the one or more client browsers navigating away from the review item.

12. The system according to claim 1, wherein:

the machine learning service;

executes an atomic claim operation that, when executed, simultaneously releases a claim by a client browser to a first review item and automatically claims a second unclaimed review item in the review queue interface.

13. The system according to claim 1, wherein:

the machine learning service;

synchronizes a client browser with other client browsers of the one or more client browsers when a viewing state of the client browser to one or more of the plurality of distinct review items changes.

14. The system according to claim 1, wherein:

the machine learning service;

pushes to a client browser of the one or more client browsers the computed state of each of a plurality of distinct review items based on a receipt of a hypertext transfer protocol (HTTP) long-polling request from the client browser.

15. The system according to claim 1, wherein:

the machine learning service;

pushes to a client browser of the one or more client browsers the computed state of each of a plurality of distinct review items based on a receipt of an hypertext transfer protocol secure (HTTPS) long-polling request from the client browser.

16. The system according to claim 1, wherein

a review item relates to an item comprising one or more of digital activity data and digital transaction data sourced from an automated decisioning workflow of the machine learning-based service without a disposal decision.