Heterogeneous graph learning-based unified network representation
A heterogeneous graph learning system generates and analyzes network implementations. The heterogeneous graph learning system includes obtaining information describing multiple network implementations including heterogeneous nodes. The heterogeneous graph learning system also includes generating a one-hop graph connecting a particular node of the heterogeneous nodes with a set of related nodes. The one-hop graph connects the particular node with the set of related nodes via corresponding edges. The heterogeneous graph learning system further includes transforming the one-hop graph into a weighted graph based on a Dynamic Meta Path Transformation (DMPT). In the DMPT, each of the corresponding edges connecting the particular node to a corresponding related node among the set of related nodes is associated with a corresponding weight.
1. A method comprising:
obtaining information describing a plurality of network implementations, the information including a heterogeneous plurality of nodes;
generating a one-hop graph connecting a particular node of the heterogeneous plurality of nodes with a set of related nodes of the heterogeneous plurality of nodes, the one-hop graph connecting the particular node with the set of related nodes via corresponding edges; and
transforming the one-hop graph into a weighted graph based on a Dynamic Meta Path Transformation (DMPT), wherein each of the corresponding edges connecting the particular node to a corresponding related node among the set of related nodes is associated with a corresponding weight.
2. The method of claim 1 , wherein the heterogeneous plurality of nodes includes nodes representing a plurality of node types selected from a group consisting of hardware nodes, software nodes, device nodes, service nodes, license nodes, event nodes, and organization nodes.
3. The method of claim 1 , wherein the DMPT comprises pruning one or more of the corresponding edges based on redundancy.
4. The method of claim 1 , wherein the DMPT comprises merging one or more nodes in the set of related nodes based on the one or more nodes having similar features.
5. The method of claim 1 , wherein the DMPT comprises determining a weighting transformation for each corresponding edge based on a node type of the corresponding related node.
6. The method of claim 1 , further comprising:
generating a plurality of weighted graphs corresponding to a plurality of nodes in the heterogeneous plurality of nodes; and
combining the plurality of weighted graphs to generate a subgraph of nodes representing a particular network implementation among the plurality of network implementations.
7. The method of claim 6 , further comprising:
generating a plurality of subgraphs representing the plurality of network implementations; and
performing soft subgraph clustering for subgraphs among the plurality of subgraphs that share one or more common nodes to adjust corresponding weights of the corresponding edges.
8. An apparatus comprising:
a memory;
a network interface configured to enable network communications; and
a processor, wherein the processor is configured to perform a method comprising:
obtaining information describing a plurality of network implementations, the information including a heterogeneous plurality of nodes;
generating a one-hop graph connecting a particular node of the heterogeneous plurality of nodes with a set of related nodes of the heterogeneous plurality of nodes, the one-hop graph connecting the particular node with the set of related nodes via corresponding edges; and
transforming the one-hop graph into a weighted graph based on a Dynamic Meta Path Transformation (DMPT), wherein each of the corresponding edges connecting the particular node to a corresponding related node among the set of related nodes is associated with a corresponding weight.
9. The apparatus of claim 8 , wherein the heterogeneous plurality of nodes includes nodes representing a plurality of node types selected from a group consisting of hardware nodes, software nodes, device nodes, service nodes, license nodes, event nodes, and organization nodes.
10. The apparatus of claim 8 , wherein the DMPT comprises pruning one or more of the corresponding edges based on redundancy.
11. The apparatus of claim 8 , wherein the DMPT comprises merging one or more nodes in the set of related nodes based on the one or more nodes having similar features.
12. The apparatus of claim 8 , wherein the DMPT comprises determining a weighting transformation for each corresponding edge based on a node type of the corresponding related node.
13. The apparatus of claim 8 , wherein the processor is further configured to perform:
generating a plurality of weighted graphs corresponding to a plurality of nodes in the heterogeneous plurality of nodes; and
combining the plurality of weighted graphs to generate a subgraph of nodes representing a particular network implementation among the plurality of network implementations.
14. The apparatus of claim 13 , wherein the processor is further configured to perform:
generating a plurality of subgraphs representing the plurality of network implementations; and
performing soft subgraph clustering for subgraphs among the plurality of subgraphs that share one or more common nodes to adjust corresponding weights of the corresponding edges.
15. One or more non-transitory computer readable storage media encoded with software comprising computer executable instructions that, when executed by a processor, cause the processor to perform a method including:
obtaining information describing a plurality of network implementations, the information including a heterogeneous plurality of nodes;
generating a one-hop graph connecting a particular node of the heterogeneous plurality of nodes with a set of related nodes of the heterogeneous plurality of nodes, the one-hop graph connecting the particular node with the set of related nodes via corresponding edges; and
transforming the one-hop graph into a weighted graph based on a Dynamic Meta Path Transformation (DMPT), wherein each of the corresponding edges connecting the particular node to a corresponding related node among the set of related nodes is associated with a corresponding weight.
16. The one or more non-transitory computer readable storage media according to claim 15 , wherein the heterogeneous plurality of nodes includes nodes representing a plurality of node types selected from a group consisting of hardware nodes, software nodes, device nodes, service nodes, license nodes, event nodes, and organization nodes.
17. The one or more non-transitory computer readable storage media according to claim 15 , wherein the DMPT comprises pruning one or more of the corresponding edges based on redundancy.
18. The one or more non-transitory computer readable storage media according to claim 17 , wherein the computer executable instructions cause the processor to further perform:
generating a plurality of weighted graphs corresponding to a plurality of nodes in the heterogeneous plurality of nodes; and
combining the plurality of weighted graphs to generate a subgraph of nodes representing a particular network implementation among the plurality of network implementations.
19. The one or more non-transitory computer readable storage media according to claim 15 , wherein the DMPT comprises merging one or more nodes in the set of related nodes based on the one or more nodes having similar features.
20. The one or more non-transitory computer readable storage media according to claim 15 , wherein the DMPT comprises determining a weighting transformation for each corresponding edge based on a node type of the corresponding related node.