Training and/or utilizing recurrent neural network model to determine subsequent source(s) for electronic resource interaction
Systems, methods, and computer readable media related to training and/or utilizing a neural network model to determine, based on a sequence of sources that each have an electronic interaction with a given electronic resource, one or more subsequent source(s) for interaction with the given electronic resource. For example, source representations of those sources can be sequentially applied (in an order that conforms to the sequence) as input to a trained recurrent neural network model, and output generated over the trained recurrent neural network model based on the applied input. The generated output can indicate, for each of a plurality of additional sources, a probability that the additional source will subsequently (e.g., next) interact with the given electronic resource. Such probabilities indicated by the output can be utilized in performance of further electronic action(s) related to the given electronic resource.
1. A method implemented by one or more processors, comprising:
determining a sequence of activated sources that each have an electronic interaction with a given electronic resource,
wherein the sequence orders the activated sources based on a chronological order of the electronic interactions with the given electronic resource by the activated sources;
generating source representations, each of the source representations being a source embedding of corresponding source features of a corresponding one of the activated sources;
applying, as input to a trained recurrent neural network model, the source representations of the activated sources and a resource representation of the given electronic resource,
wherein applying the source representations comprises applying the source representations sequentially in an order that conforms to the determined sequence of the corresponding activated sources;
generating, over the trained recurrent neural network model based on applying the input, an output; and
determining, based on the output generated over the trained recurrent neural network model, whether to provide information related to the given electronic resource; and
in response to determining, based on the output, to provide the information related to the given electronic resource:
transmitting, to one or more client devices, the information related to the given electronic resource.
2. The method of claim 1 , wherein applying the resource representation comprises:
applying the resource representation sequentially in combination with each application of the source representations.
3. The method of claim 1 , wherein the output includes at least one probability measure and wherein determining, based on the output, to provide the information related to the given electronic resource comprises determining, based on the at least one probability measure satisfying a threshold, to provide the information related to the given electronic resource.
4. The method of claim 1 , wherein transmitting, to the one or more client devices, the information related to the given electronic resource, occurs independent of receiving a query.
5. The method of claim 1 , wherein the trained recurrent neural network model comprises one or more long short-term memory units.
6. A system, comprising:
memory storing instructions;
one or more processors executing the instructions to perform a method comprising:
determining a sequence of activated sources that each have an electronic interaction with a given electronic resource,
wherein the sequence orders the activated sources based on a chronological order of the electronic interactions with the given electronic resource by the activated sources;
generating source representations, each of the source representations being a source embedding of corresponding source features of a corresponding one of the activated sources;
applying, as input to a trained recurrent neural network model, the source representations of the activated sources and a resource representation of the given electronic resource,
wherein applying the source representations comprises applying the source representations sequentially in an order that conforms to the determined sequence of the corresponding activated sources;
generating, over the trained recurrent neural network model based on applying the input, an output; and
determining, based on the output generated over the trained recurrent neural network model, whether to provide information related to the given electronic resource; and
in response to determining, based on the output, to provide the information related to the given electronic resource:
transmitting, to one or more client devices, the information related to the given electronic resource.
7. The system of claim 6 , wherein applying the resource representation comprises:
applying the resource representation sequentially in combination with each application of the source representations.
8. The system of claim 6 , wherein the output includes at least one probability measure and wherein determining, based on the output, to provide the information related to the given electronic resource comprises determining, based on the at least one probability measure satisfying a threshold, to provide the information related to the given electronic resource.
9. The system of claim 6 , wherein transmitting, to the one or more client devices, the information related to the given electronic resource, occurs independent of receiving a query.