Cross-domain user action prediction using machine learning models

A computer program product comprising a non-transitory computer-readable storage medium having instructions encoded thereon that, when executed by a processor, cause the processor to perform steps comprising:

• receiving, at an online system, a request from a client device associated with a user of the online system to access a user interface for placing an order from a source or a restaurant;
• retrieving a set of user data for the user, the set of user data comprising information describing a first set of actions in a first domain performed by the user, wherein the first domain is associated with one or more sources;
• accessing a machine-learning model trained to predict a restaurant action score indicating a likelihood of a second set of actions in a second domain performed by the user if presented with a recommendation associated with a restaurant, wherein the second domain is associated with one or more restaurants and the machine-learning model is trained by:
• gathering a set of training examples, each training example of the set of training examples comprising information describing:
• a set of previous actions in the first domain performed by each user of a plurality of users of the online system, a restaurant associated with each recommendation of one or more recommendations presented to each user of the plurality of users, and, for each recommendation of the one or more recommendations presented to each user of the plurality of users, whether a corresponding user performed one or more actions in the second domain when presented with a corresponding recommendation, and
• updating a set of parameters of the machine-learning model based at least in part on the set of training examples;
• for each candidate restaurant of a plurality of candidate restaurants associated with the online system, applying the machine-learning model to predict the restaurant action score associated with the candidate restaurant based at least in part on the set of user data for the user and a set of restaurant data for the candidate restaurant;
• selecting a set of restaurants from the plurality of candidate restaurants based at least in part on the restaurant action score associated with each candidate restaurant of the plurality of candidate restaurants;
• generating the user interface comprising a set of recommendations associated with the set of restaurants;
• and
• sending the user interface to the client device associated with the user, causing the client device to display the user interface.

A computer system comprising:

• a processor;
• and
• a non-transitory computer-readable storage medium storing instructions that, when executed by the processor, perform actions comprising:
• receiving, at an online system, a request from a client device associated with a user of the online system to access a user interface for placing an order from a source or a restaurant;
• retrieving a set of user data for the user, the set of user data comprising information describing a first set of actions in a first domain performed by the user, wherein the first domain is associated with one or more sources;
• accessing a machine-learning model trained to predict a restaurant action score indicating a likelihood of a second set of actions in a second domain performed by the user if presented with a recommendation associated with a restaurant, wherein the second domain is associated with one or more restaurants and the machine-learning model is trained by:
• gathering a set of training examples, each training example of the set of training examples comprising information describing:
• a set of previous actions in the first domain performed by each user of a plurality of users of the online system, a restaurant associated with each recommendation of one or more recommendations presented to each user of the plurality of users, and, for each recommendation of the one or more recommendations presented to each user of the plurality of users, whether a corresponding user performed one or more actions in the second domain when presented with a corresponding recommendation, and
• updating a set of parameters of the machine-learning model based at least in part on the set of training examples;
• for each candidate restaurant of a plurality of candidate restaurants associated with the online system, applying the machine-learning model to predict the restaurant action score associated with the candidate restaurant based at least in part on the set of user data for the user and a set of restaurant data for the candidate restaurant;
• selecting a set of restaurants from the plurality of candidate restaurants based at least in part on the restaurant action score associated with each candidate restaurant of the plurality of candidate restaurants;
• generating the user interface comprising a set of recommendations associated with the set of restaurants;
• and
• sending the user interface to the client device associated with the user, causing the client device to display the user interface.