System and method for obtaining recommendations using scalable cross-domain collaborative filtering

Abstract

Aspects of the present disclosure involve systems, methods, devices, and the like for presenting a recommendation. In one embodiment, a system is introduced that includes a plurality of models for obtaining a recommendation score. The recommendation score may be obtained using one or more models which can include supervised and unsupervised learning as well as a combination of user information and transactions. In another embodiment, the system is introduced that can provide a total recommendation score and recommendation generated by an ensemble model whose input can include the one or more recommendation scores previously obtained.

1 . (canceled) 
     
     
         2 . A system, comprising:
 a non-transitory memory; and   one or more hardware processors coupled to the non-transitory memory and configured to read instructions from the non-transitory memory to cause the system to perform operations comprising: 
 detecting an indication of a transaction based on an interaction by a user with a payment application associated with a merchant; 
 in response to the detecting, accessing user information associated with the user and cross-domain information representing one or more correlations between first transactions conducted with the merchant and second transactions conducted with an entity different from the merchant; 
 determining, using a first model and based on a first portion of the user information, a first recommendation score; 
 determining, using a second model based on a second portion of the user information and the cross-domain information, a second recommendation score; 
 providing the first recommendation score and the second recommendation score as inputs to a third model, wherein the third model is configured to predict, based on the first recommendation score and the second recommendation score, a correlation between the user and the entity; and 
 presenting, via a graphical user interface of the payment application, a recommendation for an item associated with the entity based on the correlation. 
   
     
     
         3 . The system of  claim 2 , wherein the item comprises a product or a service offered for sale by the entity. 
     
     
         4 . The system of  claim 2 , wherein the third model is further configured to apply a first weight to the first recommendation score and a second weight to the second recommendation score, wherein the correlation is predicted further based on the first weight and the second weight. 
     
     
         5 . The system of  claim 2 , wherein the user information represents a plurality of entities with which the user has conducted transactions in the past. 
     
     
         6 . The system of  claim 2 , wherein the user information represents a plurality of transactions conducted by the user, and wherein the plurality of transactions comprises a first transaction conducted by the user using a first payment platform, and a second transaction conducted by the user using a second payment platform. 
     
     
         7 . The system of  claim 2 , wherein the first model is configured to use a random walk algorithm to generate the first recommendation score. 
     
     
         8 . The system of  claim 2 , wherein the first model is configured to use a clustering algorithm to generate the first recommendation score. 
     
     
         9 . A method comprising:
 in response to detecting a transaction conducted between a user and a merchant via a payment application associated with the merchant, accessing user information associated with the user and cross-domain information representing one or more correlations between with the merchant and an entity different from the merchant;   determining, using a first model and based on the user information, a first recommendation score;   determining, using a second model based on the user information and the cross-domain information, a second recommendation score;   providing the first recommendation score and the second recommendation score as inputs to a third model, wherein the third model is configured to generate, based on the first recommendation score and the second recommendation score, a correlation score between the user and the entity; and   in response to determining that the correlation score exceeds a threshold, presenting, via a graphical user interface of the payment application, an item recommendation associated with an item offered for sale by the entity.   
     
     
         10 . The method of  claim 9 , wherein the payment application is implemented within a website associated with the merchant. 
     
     
         11 . The method of  claim 9 , further comprising:
 determining an indication of whether a second transaction has been conducted by the user for purchasing the item from the entity; and   training at least one of the first model, the second model, or the third model based on the indication.   
     
     
         12 . The method of  claim 9 , wherein the user information comprises at least one of a gender of the user, an age of the user, a residential address of the user, or a marital status of the user. 
     
     
         13 . The method of  claim 9 , wherein the user information represents one or more peer-to-peer transactions conducted by the user. 
     
     
         14 . The method of  claim 9 , wherein the third model comprises a decision tree model that uses the first recommendation score and the second recommendation score to generate the correlation score. 
     
     
         15 . The method of  claim 9 , wherein the third model is further configured to apply a first weight to the first recommendation score and a second weight to the second recommendation score, wherein the correlation score is generated further based on the first weight and the second weight. 
     
     
         16 . A non-transitory machine-readable medium having stored thereon machine-readable instructions executable to cause a machine to perform operations comprising:
 detecting an indication of a transaction based on one or more interactions by a user with a payment application associated with a first entity;   accessing user information associated with the user and cross-domain information associated with the first entity and a second entity;   determining based the user information, a first recommendation score;   determining, based on the user information and the cross-domain information, a second recommendation score;   providing the first recommendation score and the second recommendation score as inputs to a model, wherein the model is configured to predict, based on the first recommendation score and the second recommendation score, a correlation between the user and the second entity; and   presenting, via a graphical user interface of the payment application, a recommendation for an item associated with the second entity based on the correlation.   
     
     
         17 . The non-transitory machine-readable medium of  claim 16 , wherein the payment application corresponds to a website associated with the first entity. 
     
     
         18 . The non-transitory machine-readable medium of  claim 16 , wherein the item comprises a product or a service offered for sale by the second entity. 
     
     
         19 . The non-transitory machine-readable medium of  claim 16 , wherein the model is further configured to apply a first weight to the first recommendation score and a second weight to the second recommendation score, wherein the correlation is predicted further based on the first weight and the second weight. 
     
     
         20 . The non-transitory machine-readable medium of  claim 16 , wherein the user information represents a plurality of entities with which the user has conducted transactions in the past. 
     
     
         21 . The non-transitory machine-readable medium of  claim 9 , wherein the operations further comprise:
 determining a second indication of whether a second transaction has been conducted by the user for purchasing the item from the second entity; and   training the model based on the second indication.