System and methods for determining an object property

Abstract

A computer system for matching users with visual items comprises a user-scoring component configured to receive user data about each of the users, and process the user data so as to assign the user a set of psychological test scores that characterize the user in terms of standardized personality traits of a predefined psychological test. Ann item-scoring component uses visual appearance information of each item to assign the item a corresponding set of psychological test scores indicating expected user visual preference for the item in terms of said standardized personality traits of the psychological test. A matching component matches each user with a set of the items, by matching the user's set of psychological test scores with the corresponding sets of psychological test scores of the items, and provide a matching output indicating, to the user, the set of items with which he or she has been matched.

1 . (canceled) 
     
     
         2 . A computer system for classifying items, the computer system comprising:
 one or more hardware processors configured to implement:   a feature extractor configured to compute a domain-specific numerical feature embedding based on a combination of image data associated with an item and a textual description associated with the item;   an item classification component configured to take the domain-specific numerical feature embedding as input and output a domain-specific class for the item; and   an item scoring component configured to take the domain-specific numerical feature embedding as input and output a score vector in a personality trait space, wherein a dimensionality of the domain-specific numerical feature embedding is higher than a dimensionality of the personality trait space.   
     
     
         3 . The computer system of  claim 2 , wherein the item classification component and an embedding component of the feature extractor are trained simultaneously thereby reducing an amount of training data. 
     
     
         4 . The computer system of  claim 3 , wherein the embedding component and the item classification component are trained based on domain-specific training set using a classification loss function. 
     
     
         5 . The computer system of  claim 4 , wherein the embedding component is simultaneously trained with the item classification component to extract higher-level and domain-specific features. 
     
     
         6 . The computer system of  claim 2 , wherein the one or more hardware processors are configured to further implement a graphical user interface (GUI) component displaying a recommendation output including one or more items recommended to a user. 
     
     
         7 . The computer system of  claim 6 , wherein the one or more items are selected based at least in part on a distance between a score vector associated with each item and a set of psychological test scores of the user. 
     
     
         8 . The computer system of  claim 7 , wherein the distance is measured in the personality trait space using a distance-based search algorithm and wherein the one or more items are ordered based on the distance. 
     
     
         9 . The computer system of  claim 6 , wherein the GUI is configured to receive a user input for filtering the one or more items based on the domain-specific class associated with each item. 
     
     
         10 . The computer system of  claim 9 , wherein the score vector is dynamically adjusted based on the user input. 
     
     
         11 . The computer system of  claim 10 , wherein the GUI updates a display of the recommendation output based on the adjusted score vector. 
     
     
         12 . The computer-implemented method for classifying items, the computer-implemented method comprising:
 computing, using a feature extractor, a domain-specific numerical feature embedding based on a combination of image data associated with an item and a textual description associated with the item;   supplying the domain-specific numerical feature embedding as input to an item classification component and outputting a domain-specific class for the item; and   supplying the domain-specific numerical feature embedding as input to an item scoring component and outputting a score vector in a personality trait space, wherein a dimensionality of the domain-specific numerical feature embedding is higher than a dimensionality of the personality trait space.   
     
     
         13 . The computer-implemented method of  claim 12 , wherein the item classification component and an embedding component of the feature extractor are trained simultaneously thereby reducing an amount of training data. 
     
     
         14 . The computer-implemented method of  claim 13 , wherein the embedding component and the item classification component are trained based on domain-specific training set using on a classification loss function. 
     
     
         15 . The computer-implemented method of  claim 14 , wherein the embedding component is simultaneously trained with the item classification component to extract higher-level and domain-specific features. 
     
     
         16 . The computer-implemented method of  claim 12 , further comprising displaying, on a graphical user interface (GUI), a recommendation output including one or more items recommended to a user. 
     
     
         17 . The computer-implemented method of  claim 16 , wherein the one or more items are selected based at least in part on a distance between a score vector associated with each item and a set of psychological test scores of the user. 
     
     
         18 . The computer-implemented method of  claim 17 , wherein the distance is measured in the personality trait space using a distance-based search algorithm and wherein the one or more items are ordered based on the distance. 
     
     
         19 . The computer-implemented method of  claim 16 , wherein the GUI is configured to receive a user input for filtering the one or more items based on the domain-specific class associated with each item. 
     
     
         20 . The computer-implemented method of  claim 19 , wherein the score vector is dynamically adjusted based on the user input. 
     
     
         21 . The computer-implemented method of  claim 20 , wherein the GUI updates a display of the recommendation output based on the adjusted score vector.