Knowledge transfer system for accelerating invariant network learning

Patent metadata

Field	Value
Publication number	US-10511613-B2
Application number	US-201816055675-A
Country	US
Kind code	B2
Filing date	Aug 6, 2018
Priority date	Jan 24, 2017
Publication date	Dec 17, 2019
Grant date	Dec 17, 2019

Abstract

A computer-implemented method for implementing a knowledge transfer based model for accelerating invariant network learning is presented. The computer-implemented method includes generating an invariant network from data streams, the invariant network representing an enterprise information network including a plurality of nodes representing entities, employing a multi-relational based entity estimation model for transferring the entities from a source domain graph to a target domain graph by filtering irrelevant entities from the source domain graph, employing a reference construction model for determining differences between the source and target domain graphs, and constructing unbiased dependencies between the entities to generate a target invariant network, and outputting the generated target invariant network on a user interface of a computing device.

First claim

What is claimed is:

1. A computer-implemented method executed on a processor for implementing a knowledge transfer based model for accelerating invariant network learning, the method comprising:
generating an invariant network from data streams, the invariant network representing an enterprise information network including a plurality of nodes representing entities;
employing a multi-relational based entity estimation model for transferring the entities from a source domain graph to a target domain graph by filtering irrelevant entities from the source domain graph;
employing a reference construction model for determining differences between the source and target domain graphs, and constructing unbiased dependencies between the entities to generate a target invariant network; and
outputting the generated target invariant network on a user interface of a computing device.

2. The method of claim 1 , wherein the multi-relational based entity estimation model employs an embedding based framework to calculate relevance between pairs of the entities.

3. The method of claim 2 , wherein all the entities are represented in vector space.

4. The method of claim 3 , wherein an undirected correlation between the entities is determined in the vector space.

5. The method of claim 4 , wherein an inference technique is employed to model an optimization process as a manifold learning problem.

6. The method of claim 1 , wherein the reference construction model employs a first function to model a consistency constraint between the source and target domain graphs.

7. The method of claim 6 , wherein a second function is employed to model a smoothness constraint between a predicted invariant network and an original invariant network.

8. The method of claim 7 , wherein a unified model combines the consistency constraint and the smoothness constraint.

9. A system for implementing a knowledge transfer based model for accelerating invariant network learning, the system comprising:
a memory; and
a processor in communication with the memory, wherein the processor is configured to:
generate an invariant network from data streams, the invariant network representing an enterprise information network including a plurality of nodes representing entities;
employ a multi-relational based entity estimation model for transferring the entities from a source domain graph to a target domain graph by filtering irrelevant entities from the source domain graph;
employ a reference construction model for determining differences between the source and target domain graphs, and construct unbiased dependencies between the entities to generate a target invariant network; and
output the generated target invariant network on a user interface of a computing device.

10. The system of claim 9 , wherein the multi-relational based entity estimation model employs an embedding based framework to calculate relevance between pairs of the entities.

11. The system of claim 10 , wherein all the entities are represented in vector space.

12. The system of claim 11 , wherein an undirected correlation between the entities is determined in the vector space.

13. The system of claim 12 , wherein an inference technique is employed to model an optimization process as a manifold learning problem.

14. The system of claim 9 , wherein the reference construction model employs a first function to model a consistency constraint between the source and target domain graphs.

15. The system of claim 14 , wherein a second function is employed to model a smoothness constraint between a predicted invariant network and an original invariant network.

16. The system of claim 15 , wherein a unified model combines the consistency constraint and the smoothness constraint.

17. A non-transitory computer-readable storage medium comprising a computer-readable program for implementing a knowledge transfer based model for accelerating invariant network learning, wherein the computer-readable program when executed on a computer causes the computer to perform the steps of:
generating an invariant network from data streams, the invariant network representing an enterprise information network including a plurality of nodes representing entities;
employing a multi-relational based entity estimation model for transferring the entities from a source domain graph to a target domain graph by filtering irrelevant entities from the source domain graph;
employing a reference construction model for determining differences between the source and target domain graphs, and constructing unbiased dependencies between the entities to generate a target invariant network; and
outputting the generated target invariant network on a user interface of a computing device.

18. The non-transitory computer-readable storage medium of claim 17 , wherein the multi-relational based entity estimation model employs an embedding based framework to calculate relevance between pairs of the entities.

19. The non-transitory computer-readable storage medium of claim 18 , wherein all the entities are represented in vector space.

20. The non-transitory computer-readable storage medium of claim 19 ,
wherein an undirected correlation between the entities is determined in the vector space; and
wherein an inference technique is employed to model an optimization process as a manifold learning problem.

Assignees

• Nec Lab America Inc
• Nec Corp

Inventors

• Chen Zhengzhang
• Tang Luan
• Li Zhichun
• Luo Chen

Classifications

• H04L41/145

  involving simulating, designing, planning or modelling of a network · CPC title

• H04L43/16

  Threshold monitoring · CPC title

• H04L63/1433

  Vulnerability analysis · CPC title

• H04L41/024

  using relational databases for representation of network management data, e.g. managing via structured query language [SQL] · CPC title

• H04L41/142

  using statistical or mathematical methods · CPC title