Method and system for link prediction in large multiplex networks

Abstract

A method and a system for using a graph neural network framework to implement a link prediction in a multiplex network environment is provided. The method includes: identifying a plurality of layers of a multiplex network, each respective layer including a respective plurality of nodes; for each node included in at least a first layer, providing, by a structural node label and determining a common embedding across all of the plurality of layers and an individual embedding for each individual layer; using a k-nearest approach to select a subset of the plurality of layers for performing link prediction with respect to each layer based on the determined embeddings; and performing a link prediction by determining a respective feed-forward network with respect to each layer included in the selected subset.

What is claimed is: 
     
         1 . A method for performing a link prediction in a multiplex network environment, the method being implemented by at least one processor, the method comprising:
 identifying, by the at least one processor, a plurality of layers of a multiplex network, each respective layer of the plurality of layers including a respective plurality of nodes;   for each node included in at least a first layer from among the plurality of layers, determining, by the at least one processor, a common embedding across all of the plurality of layers and an individual embedding for each individual layer from among the plurality of layers;   using, by the at least one processor, a k-nearest approach to select a subset of the plurality of layers for performing link prediction with respect to each layer based on the determined embeddings; and   performing, by the at least one processor, a link prediction by determining a respective feed-forward network with respect to each layer included in the selected subset.   
     
     
         2 . The method of  claim 1 , further comprising providing a structural node label for each node included in at least ten layers from among the plurality of layers,
 wherein the determining of the common embedding and the individual embedding is performed for each node included in the at least ten layers from among the plurality of layers.   
     
     
         3 . The method of  claim 2 , wherein the providing of the structural node label comprises using a double-radius node labeling protocol to assign a respective label to each node. 
     
     
         4 . The method of  claim 1 , wherein a value of k is selected based on computing a nearest neighbor embedding across all of the layers included in the plurality of layers. 
     
     
         5 . The method of  claim 1 , wherein the using of the k-nearest approach to select the subset comprises randomly selecting a predetermined number of layers for performing the link prediction for each layer. 
     
     
         6 . The method of  claim 1 , wherein the using of the k-nearest approach to select the subset comprises:
 selecting a predetermined number of layers from among the plurality of layers;   determining, for each of the predetermined number of layers, a respective count of a number of selections by a k-nearest layer; and   selecting the subset for performing the link prediction based on the determined counts for all layers included in the predetermined number of layers.   
     
     
         7 . The method of  claim 1 , wherein the performing of the link prediction for a particular layer comprises using a self-attention mechanism to combine information that relates to at least two layers that are different from the particular layer with the determined embeddings in order to perform the link prediction. 
     
     
         8 . A computing apparatus for performing a link prediction in a multiplex network environment, the computing apparatus comprising:
 a processor;   a memory; and   a communication interface coupled to each of the processor and the memory,   wherein the processor is configured to:
 identify a plurality of layers of a multiplex network, each respective layer of the plurality of layers including a respective plurality of nodes; 
 for each node included in at least a first layer from among the plurality of layers, determine a common embedding across all of the plurality of layers and an individual embedding for each individual layer from among the plurality of layers; 
 use a k-nearest approach to select a subset of the plurality of layers for performing link prediction with respect to each layer based on the determined embeddings; and 
 perform a link prediction by determining a respective feed-forward network with respect to each layer included in the selected subset. 
   
     
     
         9 . The computing apparatus of  claim 8 , wherein the processor is further configured to provide a structural node label and to determine of the common embedding and the individual embedding for each node included in at least ten layers from among the plurality of layers. 
     
     
         10 . The computing apparatus of  claim 9 , wherein the processor is further configured to provide the structural node label by using a double-radius node labeling protocol to assign a respective label to each node. 
     
     
         11 . The computing apparatus of  claim 8 , wherein the processor is further configured to select a value of k based on computing a nearest neighbor embedding across all of the layers included in the plurality of layers. 
     
     
         12 . The computing apparatus of  claim 8 , wherein the processor is further configured to use the k-nearest approach to select the subset by randomly selecting a predetermined number of layers for performing the link prediction for each layer. 
     
     
         13 . The computing apparatus of  claim 8 , wherein the processor is further configured to use the k-nearest approach to select the subset by:
 selecting a predetermined number of layers from among the plurality of layers;   determining, for each of the predetermined number of layers, a respective count of a number of selections by a k-nearest layer; and   selecting the subset for performing the link prediction based on the determined counts for all layers included in the predetermined number of layers.   
     
     
         14 . The computing apparatus of  claim 8 , wherein the processor is further configured to perform the link prediction for a particular layer by using a self-attention mechanism to combine information that relates to at least two layers that are different from the particular layer with the determined embeddings in order to perform the link prediction. 
     
     
         15 . A non-transitory computer readable storage medium storing instructions for performing a link prediction in a multiplex network environment, the storage medium comprising executable code which, when executed by a processor, causes the processor to:
 identify a plurality of layers of a multiplex network, each respective layer of the plurality of layers including a respective plurality of nodes;   for each node included in at least a first layer from among the plurality of layers, determine a common embedding across all of the plurality of layers and an individual embedding for each individual layer from among the plurality of layers;   use a k-nearest approach to select a subset of the plurality of layers for performing link prediction with respect to each layer based on the determined embeddings; and   perform a link prediction by determining a respective feed-forward network with respect to each layer included in the selected subset.   
     
     
         16 . The storage medium of  claim 15 , wherein the executable code is further configured to cause the processor to provide a structural node label and to determine the common embedding and the individual embedding for each node included in at least ten layers from among the plurality of layers. 
     
     
         17 . The storage medium of  claim 16 , wherein the executable code is further configured to cause the processor to provide the structural node label by using a double-radius node labeling protocol to assign a respective label to each node. 
     
     
         18 . The storage medium of  claim 15 , wherein the executable code is further configured to cause the processor to select a value of k based on computing a nearest neighbor embedding across all of the layers included in the plurality of layers. 
     
     
         19 . The storage medium of  claim 15 , wherein the executable code is further configured to cause the processor to use the k-nearest approach to select the subset by randomly selecting a predetermined number of layers for performing the link prediction for each layer. 
     
     
         20 . The storage medium of  claim 15 , wherein the executable code is further configured to cause the processor to use the k-nearest approach to select the subset by:
 selecting a predetermined number of layers from among the plurality of layers;   determining, for each of the predetermined number of layers, a respective count of a number of selections by a k-nearest layer; and   selecting the subset for performing the link prediction based on the determined counts for all layers included in the predetermined number of layers.