Systems, methods, and devices for generating recommendations of unique items

1. A computer-implemented method for generating user specific recommendations of unique items, the computer-implemented method comprising:

monitoring, by a computer system over a computer network, a plurality of user devices to detect interactions by a user with unique item listings presented by at least one electronic listing system, wherein the unique item listings are associated with unique items each comprising a plurality of features including at least a condition and a location;

generating, by the computer system, a user specific scoring model using historical user preference data derived from the monitoring;

identifying, by the computer system, a selected unique item selected by the user, wherein the selected unique item comprises a unique item presented by the at least one electronic listing system, the selected unique item comprising a plurality of features including at least a condition and a location;

determining, by the computer system, differences between the plurality of features of the selected unique item and a plurality of features of each of a plurality of alternative unique items;

generating, by the computer system, a recommendation score for each of the plurality of alternative unique items, wherein the recommendation score indicates a predicted level of preference by the user;

wherein generating the recommendation score comprises applying, by the computer system, the user specific scoring model to the determined differences;

generating, by the computer system, a list of recommended alternative unique items that are predicted to be preferred by the user, wherein the list comprises at least a portion of the plurality of alternative unique items sorted by their respective recommendation scores;

rendering an interactive electronic interface of the at least one electronic listing system, wherein the interactive electronic interface presents the unique item listings and enables the user to search for and interact with the unique item listings;

detecting additional interactions by the user with unique item listings presented by the rendered interactive electronic interface;

regenerating, by the computer system, the user specific scoring model based at least partially on updated historical user preference data that takes into account the additional interactions;

regenerating, by the computer system, recommendation scores for at least a portion of the plurality of alternative unique items, using the regenerated user specific scoring model; and

responsive to the regenerating of recommendation scores, re-rendering the interactive electronic interface to present at least a subset of the plurality of alternative unique items sorted by the regenerated recommendation scores;

wherein the computer system comprises a computer processor and electronic memory.

2. The computer-implemented method of claim 1 , wherein the user specific scoring model is configured to enable generation of recommendation scores for alternative unique items that are not included in the historical user preference data derived from the monitoring.

3. The computer-implemented method of claim 1 , further comprising:

identifying, by the computer system, based on the monitored interactions, a preferred unique item, wherein the preferred unique item is an item presented by the at least one electronic listing system which the user has interacted with in a manner indicating the user prefers the preferred unique item over at least one other unique item; and

updating the historical user preference data with an indication of the preferred unique item.

4. The computer-implemented method of claim 3 , further comprising:

updating the user specific scoring model, wherein updating the user specific scoring model comprises:

identifying, by the computer system, a preferred feature of the preferred unique item that is not a feature of the other unique item; and

updating the user specific scoring model such that generating a recommendation score will return a higher recommendation score for a unique item having the preferred feature.

5. The computer-implemented method of claim 1 , wherein the generating of the user specific scoring model comprises at least:

creating training data comprising a plurality of pairs of unique items that were detected as interacted with by the user, each of the plurality of pairs of unique items comprising difference data indicative of differences between a plurality of features including at least a condition and a location;

labeling each of the plurality of pairs with a preference indicator associated with one of the unique items of each pair; and

inputting the training data into a supervised learning algorithm to generate a function that can output a predicted relative level of preference for an alternative unique item based on differences between condition and location features of the alternative unique item and a selected unique item.

6. The computer-implemented method of claim 1 , wherein the computer system comprises the at least one electronic listing system.

7. A computer readable, non-transitory storage medium having a computer program stored thereon for causing a suitably programmed computer system to process by one or more processors computer-program code by performing a method for generating user specific recommendations of unique items when the computer program is executed on the suitably programmed computer system, the method comprising:

monitoring, by a computer system over a computer network, a plurality of user devices to detect interactions by a user with unique item listings presented by at least one electronic listing system, wherein the unique item listings are associated with unique items each comprising a plurality of features including at least a condition and a location;

generating, by the computer system, a user specific scoring model using historical user preference data derived from the monitoring;

identifying, by the computer system, a selected unique item selected by the user, wherein the selected unique item comprises a unique item presented by the at least one electronic listing system, the selected unique item comprising a plurality of features including at least a condition and a location;

determining, by the computer system, differences between the plurality of features of the selected unique item and a plurality of features of each of a plurality of alternative unique items;

generating, by the computer system, a recommendation score for each of the plurality of alternative unique items, wherein the recommendation score indicates a predicted level of preference by the user;

wherein generating the recommendation score comprises applying, by the computer system, the user specific scoring model to the determined differences;

generating, by the computer system, a list of recommended alternative unique items that are predicted to be preferred by the user, wherein the list comprises at least a portion of the plurality of alternative unique items sorted by their respective recommendation scores;

rendering an interactive electronic interface of the at least one electronic listing system, wherein the interactive electronic interface presents the unique item listings and enables the user to search for and interact with the unique item listings;

detecting additional interactions by the user with unique item listings presented by the rendered interactive electronic interface;

regenerating, by the computer system, the user specific scoring model based at least partially on updated historical user preference data that takes into account the additional interactions;

regenerating, by the computer system, recommendation scores for at least a portion of the plurality of alternative unique items, using the regenerated user specific scoring model; and

responsive to the regenerating of recommendation scores, re-rendering the interactive electronic interface to present at least a subset of the plurality of alternative unique items sorted by the regenerated recommendation scores;

wherein the computer system comprises a computer processor and electronic memory.

8. The computer readable, non-transitory storage medium of claim 7 , wherein the user specific scoring model is configured to enable generation of recommendation scores for alternative unique items that are not included in the historical user preference data derived from the monitoring.

9. The computer readable, non-transitory storage medium of claim 7 , the method further comprising:

identifying, by the computer system, based on the monitored interactions, a preferred unique item, wherein the preferred unique item is an item presented by the at least one electronic listing system which the user has interacted with in a manner indicating the user prefers the preferred unique item over at least one other unique item; and

updating the historical user preference data with an indication of the preferred unique item.

10. The computer readable, non-transitory storage medium of claim 9 , the method further comprising:

updating the user specific scoring model, wherein updating the user specific scoring model comprises:

identifying, by the computer system, a preferred feature of the preferred unique item that is not a feature of the other unique item; and

updating the user specific scoring model such that generating a recommendation score will return a higher recommendation score for a unique item having the preferred feature.

11. The computer readable, non-transitory storage medium of claim 7 , wherein the generating of the user specific scoring model comprises at least:

creating training data comprising a plurality of pairs of unique items that were detected as interacted with by the user, each of the plurality of pairs of unique items comprising difference data indicative of differences between a plurality of features including at least a condition and a location;

labeling each of the plurality of pairs with a preference indicator associated with one of the unique items of each pair; and

inputting the training data into a supervised learning algorithm to generate a function that can output a predicted relative level of preference for an alternative unique item based on differences between condition and location features of the alternative unique item and a selected unique item.

12. The computer readable, non-transitory storage medium of claim 7 , wherein the computer system comprises the at least one electronic listing system.