Machine learning-based techniques for representing computing processes as vectors

Patent metadata

Field	Value
Publication number	US-11645539-B2
Application number	US-201916518808-A
Country	US
Kind code	B2
Filing date	Jul 22, 2019
Priority date	Jul 22, 2019
Publication date	May 9, 2023
Grant date	May 9, 2023

Abstract

Machine learning-based techniques for representing computing processes as vectors are provided. In one set of embodiments, a computer system can receive a name of a computing process and context information pertaining to the computing process. The computer system can further train a neural network based on the name and the context information, where the training results in determination of weight values for one or more hidden layers of the neural network. The computer system can then generate, based on the weight values, a vector representation of the computing process that encodes the context information and can perform one or more analyses using the vector representation.

First claim

What is claimed is:

1. A method comprising:
receiving, by a computer system, a name of a computing process and context information pertaining to the computing process;
training, by the computer system, a neural network based on the name and the context information, wherein the training results in determination of weight values for one or more hidden layers of the neural network, and wherein the training comprises:
creating a one-hot-encoded vector for the computing process;
creating one-hot-encoded vectors for context-related objects of the computing process, the context-related objects being determined from the context information;
setting the one-hot-encoded vector for the computing process as an input of the neural network;
setting the one-hot-encoded vectors for the context-related objects as outputs of the neural network; and
training the neural network to determine the weight values for the one or more hidden layers in a manner that predicts the outputs from the input;

generating, by the computer system, a vector representation of the computing process based on the weight values, the vector representation encoding the context information; and
performing, by the computer system, one or more analyses using the vector representation of the computing process.

2. The method of claim 1 wherein the context information comprises a relationship between the computing process and one or more other computing processes.

3. The method of claim 1 wherein the context information comprises a relationship between the computing process and one or more features of a machine on which the computing process runs.

4. The method of claim 1 wherein the context information comprises information regarding one or more functions performed by the computing process during its runtime.

5. The method of claim 1 wherein the context-related objects are other computing processes that are determined to co-occur with the computing process.

6. The method of claim 1 wherein the one or more analyses include determining whether the computing process is similar to one or more other computing processes by calculating similarity scores between the vector representation of the computing process and vector representations of the one or more other computing processes.

7. A non-transitory computer readable storage medium having stored thereon program code executable by a computer system, the program code embodying a method comprising:
receiving a name of a computing process and context information pertaining to the computing process;
training a neural network based on the name and the context information, wherein the training results in determination of weight values for one or more hidden layers of the neural network, and wherein the training comprises:
creating a one-hot-encoded vector for the computing process;
creating one-hot-encoded vectors for context-related objects of the computing process, the context-related objects being determined from the context information;
setting the one-hot-encoded vector for the computing process as an input of the neural network;
setting the one-hot-encoded vectors for the context-related objects as outputs of the neural network; and
training the neural network to determine the weight values for the one or more hidden layers in a manner that predicts the outputs from the input;

generating a vector representation of the computing process based on the weight values, the vector representation encoding the context information; and
performing one or more analyses using the vector representation of the computing process.

8. The non-transitory computer readable storage medium of claim 7 wherein the context information comprises a relationship between the computing process and one or more other computing processes.

9. The non-transitory computer readable storage medium of claim 7 wherein the context information comprises a relationship between the computing process and one or more features of a machine on which the computing process runs.

10. The non-transitory computer readable storage medium of claim 7 wherein the context information comprises information regarding one or more functions performed by the computing process during its runtime.

11. The non-transitory computer readable storage medium of claim 7 wherein the context-related objects are other computing processes that are determined to co-occur with the computing process.

12. The non-transitory computer readable storage medium of claim 7 wherein the one or more analyses include determining whether the computing process is similar to one or more other computing processes by calculating similarity scores between the vector representation of the computing process and vector representations of the one or more other computing processes.

13. A computer system comprising:
a processor;
a neural network; and
a non-transitory computer readable medium having stored thereon program code that, when run, causes the processor to:
receive a name of a computing process and context information pertaining to the computing process;
train the neural network based on the name and the context information, wherein the training results in determination of weight values for one or more hidden layers of the neural network, and wherein the training comprises:
creating a one-hot-encoded vector for the computing process;
creating one-hot-encoded vectors for context-related objects of the computing process, the context-related objects being determined from the context information;
setting the one-hot-encoded vector for the computing process as an input of the neural network;
setting the one-hot-encoded vectors for the context-related objects as outputs of the neural network; and
training the neural network to determine the weight values for the one or more hidden layers in a manner that predicts the outputs from the input;

generate a vector representation of the computing process based on the weight values, the vector representation encoding the context information; and
perform one or more analyses using the vector representation of the computing process.

14. The computer system of claim 13 wherein the context information comprises a relationship between the computing process and one or more other computing processes.

15. The computer system of claim 13 wherein the context information comprises a relationship between the computing process and one or more features of a machine on which the computing process runs.

16. The computer system of claim 13 wherein the context information comprises information regarding one or more functions performed by the computing process during its runtime.

17. The computer system of claim 13 wherein the context-related objects are other computing processes that are determined to co-occur with the computing process.

18. The computer system of claim 13 wherein the one or more analyses include determining whether the computing process is similar to one or more other computing processes by calculating similarity scores between the vector representation of the computing process and vector representations of the one or more other computing processes.

Assignees

• Vmware Inc

Inventors

• Zan Bin
• Mo Zhen
• Akkineni Vamsi
• Ganti Vijay

Classifications

• G06N3/08

  Learning methods · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/045

  Combinations of networks · CPC title

• G06F17/16

  Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} · CPC title