Relational network inferencing using random traversal paths of a graph representation

A system comprising a computing device having a processor and a memory having executable instructions stored thereon, which, when executed, cause the processor to perform a method of:

• generating, by an application server, for a graph representation, a plurality of traversal paths including:
• a starting vertex having an initial probability value;
• and
• a random hop selected based on a weight of an edge and a total weight of selectable edges for the random hop;
• generating, by the application server, one or more entity features comprising a probability that an entity represented by a particular vertex will be impacted by an event associated with the starting vertex based on the initial probability value and a proportion of the plurality of traversal paths that includes the particular vertex, wherein the entity is a company or a person, and wherein the application server generates the one or more entity features based on a bounded computational cost graph traversal process that reduces computing resource utilization through use of a constant determined based on a number of hops;
• providing, by the application server, the one or more entity features as inputs to a machine learning model that has been trained through an iterative model optimization process based on features of the graph representation;
• generating, by the machine learning model based on the inputs, an output indicating a probability of a corresponding event relating to the entity;
• and
• performing, by the application server, an action related to the entity based on the output from the machine learning model.

A non-transitory computer-readable medium comprising instructions that, when executed by a processor, cause the processor to perform a method for determining a graph feature, the method comprising:

• generating, by an application server, for a graph representation, a plurality of traversal paths including:
• a starting vertex having an initial probability value;
• and
• a random hop selected based on a weight of an edge and a total weight of selectable edges for the random hop;
• generating, by the application server, one or more entity features comprising a probability that an entity represented by a particular vertex will be impacted by an event associated with the starting vertex based on the initial probability value and a proportion of the plurality of traversal paths that includes the particular vertex, wherein the entity is a company or a person, and wherein the application server generates the one or more entity features based on a bounded computational cost graph traversal process that reduces computing resource utilization through use of a constant determined based on a number of hops;
• providing, by the application server, the one or more entity features as inputs to a machine learning model that has been trained through an iterative model optimization process based on features of the graph representation;
• generating, by the machine learning model based on the inputs, an output indicating a probability of a corresponding event relating to the entity;
• and
• performing, by the application server, an action related to the entity based on the output from the machine learning model.