Granted Patent B2

US 12,314,821 · App. 17/661,821 · Granted May 27, 2025

Inventors: Netanel Maman (Mazkeret Batya, IL); Samuel Elgozi (Gan Yavne, IL); Tomer Semo (Tel Aviv-Jaffa, IL); Tomer Roizman (Tel Aviv-Jaffa, IL); Ofir Manzur (Tel Aviv-Jaffa, IL)

Assignee: Akamai Technologies, Inc.

G06N20/00G06F9/543G06F21/577G06F2221/034

Abstract

Claims (35)

1. A method for active parameter identification, comprising:

applying a machine learning model to features extracted from each of at least one request to a computing interface, wherein the machine learning model is trained per value using a training set including a plurality of training values of a plurality of training requests, wherein the machine learning model is trained to output an indicator as to whether each portion of a request containing a respective value indicates a parameter when applied to the request; and

identifying at least one parameter-indicating portion of each request to the computing interface based on outputs of the machine learning model.

2. The method of claim 1 , further comprising:

creating a specification for the computing interface based on the identified at least one parameter-indicating portion of each request to the computing interface.

3. The method of claim 2 , wherein each of the at least one request to the computing interface is an example request having an entropy above a threshold.

4. The method of claim 3 , wherein multiple instances of each of the at least one request to the computing interface is included in traffic to the computing interface.

5. The method of claim 2 , further comprising:

detecting at least one vulnerability of the computing interface using the specification created for the computing interface.

6. The method of claim 5 , wherein the at least one request to the computing interface is at least one first request, wherein detecting the at least one vulnerability further comprises:

consuming the computing interface by sending at least one second request based on the identified at least one parameter-indicating portion of each first request to the computing interface.

7. The method of claim 6 , further comprising:

determining at least one value to be included in the at least one second request such that each of the at least one value corresponds to one of the at least one parameter-indicating portion of one of the at least one first request to the computing interface.

8. The method of claim 7 , wherein the at least one value is determined based on at least one dependency of the computing interface.

9. The method of claim 1 , wherein the computing interface is an application programming interface.

10. A non-transitory computer readable medium having stored thereon instructions for causing a processing circuitry to execute a process, the process comprising:

applying a machine learning model to features extracted from each of at least one request to a computing interface, wherein the machine learning model is trained per value using a training set including a plurality of training values of a plurality of training requests, wherein the machine learning model is trained to output an indicator as to whether each portion of a request containing a respective value indicates a parameter when applied to the request; and

identifying at least one parameter-indicating portion of each request to the computing interface based on outputs of the machine learning model.

11. A system for active parameter identification, comprising:

a processing circuitry; and

a memory, the memory containing instructions that, when executed by the processing circuitry, configure the system to:

apply a machine learning model to features extracted from each of at least one request to a computing interface, wherein the machine learning model is trained per value using a training set including a plurality of training values of a plurality of training requests, wherein the machine learning model is trained to output an indicator as to whether each portion of a request containing a respective value indicates a parameter when applied to the request; and

identify at least one parameter-indicating portion of each request to the computing interface based on outputs of the machine learning model.

12. The system of claim 11 , wherein the system is further configured to:

create a specification for the computing interface based on the identified at least one parameter-indicating portion of each request to the computing interface.

13. The system of claim 12 , wherein each of the at least one request to the computing interface is an example request having an entropy above a threshold.

14. The system of claim 13 , wherein multiple instances of each of the at least one request to the computing interface is included in traffic to the computing interface.

15. The system of claim 12 , wherein the system is further configured to:

detect at least one vulnerability of the computing interface using the specification created for the computing interface.

16. The system of claim 15 , wherein the at least one request to the computing interface is at least one first request, wherein the system is further configured to:

consume the computing interface by sending at least one second request based on the identified at least one parameter-indicating portion of each first request to the computing interface.

17. The system of claim 16 , wherein the system is further configured to:

determine at least one value to be included in the at least one second request such that each of the at least one value corresponds to one of the at least one parameter-indicating portion of one of the at least one first request to the computing interface.

18. The system of claim 17 , wherein the at least one value is determined based on at least one dependency of the computing interface.

19. The system of claim 11 , wherein the computing interface is an application programming interface.

Assignments (2)

Continuity (1)

References Cited (18)