Personal Taste Assessment Method and System

Abstract

A personal taste assessment system recommends a consumable or other item. The system accesses a preference model for the representing associations between a person's ratings of items and a set of item characteristics. The system uses the preference model and an item database that contains characteristic values for a group of items to identify a candidate item having characteristics that correspond to traits that the person found to be appealing, and which may be chemically sensed. The system then causes an electronic device to output an identification of the candidate item.

1 . A method for recommending an item, comprising:
 by one or more processors:
 accessing a user profile for a person, wherein the user profile comprises a preference model representing associations between the person's ratings of items and a plurality of item characteristics; 
 accessing a database of characteristic values for a plurality of candidate items; 
 identifying a candidate item in the database having characteristic values whose properties match traits associated with items that the person rated and found to be appealing; 
 retrieving one or more characteristics of the identified candidate item, wherein the characteristics comprise data that may be chemically sensed; 
 processing the retrieved characteristics with the user profile to generate a predicted rating for the identified candidate of interest for the person by:
 determining whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference in the preference model, and 
 based on whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference, generating the predicted rating as a prediction of how likely the person is to like or dislike the identified candidate item; and 
 
 causing the electronic device to output the predicted rating to the person. 
   
     
     
         2 . The method of  claim 1 , wherein identifying the candidate item in the database also comprises constraining the identified candidate item to an item having one or more characteristics that satisfy a user-supplied criterion. 
     
     
         3 . The method of  claim 2 , wherein the user-supplied criterion comprises a color, a price range, a maximum price or a type. 
     
     
         4 . The method of  claim 1 , wherein identifying the candidate item in the database also comprises constraining the identified candidate item to one having one or more characteristics that satisfy a situational criterion. 
     
     
         5 . The method of  claim 4 , wherein the situational criterion comprises available inventory at a retailer. 
     
     
         6 . The method of  claim 1 , wherein identifying the candidate item in the database also comprises not selecting any item having one or more characteristics for which the person has exhibited a negative preference. 
     
     
         7 . The method of  claim 1 , wherein the characteristic values of the identified consumable item in the database are associated with at least one of the following senses: taste, smell and touch. 
     
     
         8 . The method of  claim 1 , further comprising:
 receiving a context for the person's experience with one or more of items that the person has rated; and   identifying the candidate item in the database comprises identifying an item having characteristics consistent with preferences of the person under conditions specified by a contextual criterion that corresponds to the context.   
     
     
         9 . The method of  claim 1 , further comprising:
 accessing the database to identify at least one additional item that exhibits one or more characteristics that, if the additional item is rated by the person, will improve the preference model for the person; and   generating a suggestion that the person rate the additional item.   
     
     
         10 . A system for generating a recommendation for a consumable item, comprising:
 one or more processors:   a database of characteristic values for a plurality of candidate items;   a computer-readable memory storing a user profile for a person, wherein the user profile comprises a preference model representing associations between the person's ratings of items and a plurality of item characteristics; and   a computer-readable memory containing programming instructions that are configured to, when executed, cause the one or more processors to:
 access the user profile and the database of characteristic values; 
 identify a candidate item in the database having characteristic values whose properties match traits associated with items that the person rated and found to be appealing; 
 retrieve one or more characteristics of the identified candidate item, wherein the characteristics comprise data that may be chemically sensed; 
 process the retrieved characteristics with the user profile to generate a predicted rating for the identified candidate of interest for the person by:
 determining whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference in the preference model, and 
 based on whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference, generating the predicted rating as a prediction of how likely the person is to like or dislike the identified candidate item; and 
 
 cause the electronic device to output the predicted rating to the person. 
   
     
     
         11 . The system of  claim 10 , wherein the instructions to identify a candidate item also comprise instructions to constrain the candidate item to one having one or more characteristics that satisfy a user-supplied criterion. 
     
     
         12 . The system of  claim 11 , wherein the user-supplied criterion comprises a color, a price range, a maximum price or a type. 
     
     
         13 . The system of  claim 10 , wherein the instructions to identify a candidate item also comprise instructions to constrain the candidate item to one having one or more characteristics that satisfy a situational criterion. 
     
     
         14 . The system of  claim 13 , wherein the situational criterion comprises available inventory at a retailer. 
     
     
         15 . The system of  claim 10 , wherein instructions to identify the candidate item also comprise instructions to not select any candidate item having one or more characteristics for which the person has exhibited a negative preference. 
     
     
         16 . The system of  claim 10 , wherein the characteristic values of the identified candidate item comprise values of traits that are associated with at least one of the following senses: taste, smell and touch. 
     
     
         17 . The system of  claim 10 , further comprising instructions to:
 generate a predicted rating for the identified candidate item; and   cause the electronic device to output the predicted rating with the identification of the identified candidate item.   
     
     
         18 . The system of  claim 10 , further comprising instructions to:
 receive a context for the person's experience with one or more of the rated consumable items; and   when identifying the candidate item, identifying an item having characteristics consistent with preferences of the person under conditions specified by a contextual criterion that corresponds to the context.   
     
     
         19 . The system of  claim 11 , further comprising instructions to:
 access the item database to identify at least one additional item that exhibits one or more characteristics that, if the additional item is rated by the person, will improve the preference model for the person; and   generate a suggestion that the person rate the additional item.   
     
     
         20 . A computer program product comprising a computer-readable medium containing programming instructions that are configured to instruct one or more processors to:
 access a user profile for a person, wherein the user profile comprises a preference model representing associations between the person's ratings of items and a plurality of item characteristics;   access a database of characteristic values for a plurality of candidate items;   identify a candidate item in the database having characteristic values whose properties match traits associated with items that the person rated and found to be appealing;   retrieve one or more characteristics of the identified candidate item, wherein the characteristics comprise data that may be chemically sensed;   process the retrieved characteristics with the user profile to generate a predicted rating for the identified candidate of interest for the person by:   determining whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference in the preference model, and   based on whether the retrieved characteristics of the identified candidate item correspond to one or more indicators of preference, generating the predicted rating as a prediction of how likely the person is to like or dislike the identified candidate item; and   cause an electronic device to output the predicted rating to the person.