Facilitated comparison of gemstones

Abstract

Methods and systems are provided for facilitating comparison of gemstones having multiple attributes that characterize the gemstones. A facilitated comparison system receives from a user a selection of a gemstone in a collection of the gemstones. Each gemstone in the collection is associated with a selling index indicating a probability of the gemstone being bought by a customer. For each of a predetermined set of the attributes, the system determines one or more comparable ranges. The system identifies alternate gemstones to recommend using the comparable ranges for the attributes and the selling indices of the gemstones. The system displays to the user the selected gemstone and one or more of the alternate gemstones for comparison.

I/We claim: 
     
         1 . A non-transitory computer-readable medium carrying instructions that, when executed by a computing system, cause the computing system to perform operations to identify alternative gemstones for comparison with a selected gemstone, the operations comprising:
 maintaining a dataset of historical customer interaction data with gemstones, the dataset including information about gemstones viewed and purchased by customers, the gemstones characterized by two or more attributes selected from a price, a weight, a cut, a color, or a clarity;   selecting a time period of historical customer interaction data;   applying machine learning to the selected time period of historical customer interaction data to generate a predictive model calculator of a probability of a gemstone being bought by a customer based on two or more attributes of the gemstone and past customer interaction data for gemstones having similar attributes;   receiving a selection of a gemstone in a set of gemstones,
 wherein each gemstone in the set of gemstones is associated with multiple attributes characterizing the gemstone and a selling probability calculated using the predictive model calculator, the selling probability indicating a likelihood that the gemstone will be purchased; 
   determining comparable ranges for a portion of the multiple attributes characterizing the selected gemstone;   identifying alternative gemstones based on the determined comparable ranges; and   causing display of the selected gemstone and at least one alternative gemstone of the identified alternative gemstones based in part on a selling probability for the at least one alternative gemstone.   
     
     
         2 . The computer-readable medium of  claim 1 , wherein the historical customer interaction data additionally includes at least one of gemstones added to a shopping cart or gemstones viewed by customers who bought a gemstone in a given time period. 
     
     
         3 . The computer-readable medium of  claim 1 , wherein a comparable range for an attribute is determined using a dynamic threshold based on a value of the attribute. 
     
     
         4 . The computer-readable medium of  claim 1 , wherein the selected gemstone and the at least one alternative gemstone of the identified alternative gemstones are displayed in a grid view comprising a horizontal axis representing a first gemstone attribute and a vertical axis representing a second gemstone attribute. 
     
     
         5 . The computer-readable medium of  claim 4 , wherein the first attribute comprises a gemstone quality and the second attribute comprises a gemstone price range. 
     
     
         6 . The computer-readable medium of  claim 4 , wherein a portion of the grid is unpopulated. 
     
     
         7 . The computer-readable medium of  claim 4 , wherein each box in the grid is associated with a descriptive label that describes gemstones that satisfy criteria of the box. 
     
     
         8 . The computer-readable medium of  claim 1 , wherein causing display of the selected gemstone and the at least one alternative gemstone of the identified alternative gemstones includes causing display of at least one visual cue associated with an attribute of the at least one alternative gemstone, the visual cue indicating a comparison between the attribute of the at least one alternative gemstone and a corresponding attribute of the selected gemstone. 
     
     
         9 . The computer-readable medium of  claim 1 , wherein the predictive model calculator is selected from multiple predictive models based on results of cross-validation of the predictive models. 
     
     
         10 . The computer-readable medium of  claim 1 , wherein identifying the alternative gemstones is further based on expected delivery dates. 
     
     
         11 . A computer-implemented method to identify alternative gemstones for comparison with a selected gemstone, the method comprising:
 maintaining a dataset of historical customer interaction data with gemstones, the dataset including information about gemstones viewed and purchased by customers, the gemstones characterized by two or more attributes selected from a price, a weight, a cut, a color, or a clarity;   selecting a time period of historical customer interaction data;   applying machine learning to the selected time period of historical customer interaction data to generate a predictive model calculator of a probability of a gemstone being bought by a customer based on two or more attributes of the gemstone and past customer interaction data for gemstones having similar attributes;   receiving a selection of a gemstone in a set of gemstones,
 wherein each gemstone in the set of gemstones is associated with multiple attributes characterizing the gemstone and a selling probability calculated using the predictive model calculator, the selling probability indicating a likelihood that the gemstone will be purchased; 
   determining comparable ranges for a portion of the multiple attributes characterizing the selected gemstone;   identifying alternative gemstones based on the determined comparable ranges; and   causing display of the selected gemstone and at least one alternative gemstone of the identified alternative gemstones based in part on a selling probability for the at least one alternative gemstone.   
     
     
         12 . The computer-implemented method of  claim 11 , wherein the historical customer interaction data additionally includes at least one of gemstones added to a shopping cart or gemstones viewed by customers who bought a gemstone in a given time period. 
     
     
         13 . The computer-implemented method of  claim 11 , wherein a comparable range for an attribute is determined using a dynamic threshold based on a value of the attribute. 
     
     
         14 . The computer-implemented method of  claim 11 , wherein the selected gemstone and the at least one alternative gemstone of the identified alternative gemstones are displayed in a grid view comprising a horizontal axis representing a first gemstone attribute and a vertical axis representing a second gemstone attribute. 
     
     
         15 . The computer-implemented method of  claim 14 , wherein the first attribute comprises a gemstone quality and the second attribute comprises a gemstone price range. 
     
     
         16 . The computer-implemented method of  claim 14 , wherein a portion of the grid is unpopulated. 
     
     
         17 . The computer-implemented method of  claim 14 , wherein each box in the grid is associated with a descriptive label that describes gemstones that satisfy criteria of the box. 
     
     
         18 . The computer-implemented method of  claim 11 , wherein causing display of the selected gemstone and the at least one alternative gemstone of the identified alternative gemstones includes causing display of at least one visual cue associated with an attribute of the at least one alternative gemstone, the visual cue indicating a comparison between the attribute of the at least one alternative gemstone and a corresponding attribute of the selected gemstone. 
     
     
         19 . The computer-implemented method of  claim 11 , wherein the predictive model calculator is selected from multiple predictive models based on results of cross-validation of the predictive models. 
     
     
         20 . The computer-implemented method of  claim 11 , wherein identifying the alternative gemstones is further based on expected delivery dates.