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

• US 10,007,946 B1
• Filed: 07/09/2015
• Issued: 06/26/2018
• Est. Priority Date: 03/07/2013
• Status: Active Grant

• Current Assignee: VAST.com, Vast.com, Inc. (Vroom, Inc.)
• Original Assignee: Vast.com, Inc. (Vroom, Inc.)
• Inventors: Levy, Joshua Howard, Feldman, Lauri Janet, Dove, Andrew Philip
• Primary Examiner(s): Chen, Alan
• Application Number:
  US14/795,809
• Time in Patent Office: 1,083 Days
• Field of Search: None
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06Q 30/0631 Item recommendations

Abstract

The disclosure herein provides methods, systems, and devices for measuring similarity of and generating recommendations for unique items. A recommendation system for generating recommendations of alternative unique items comprises an items information database, a penalty computation engine, a recommendation compilation engine, and one or more computers, wherein the penalty computation engine comprises a customizations filter, a condition filter, and a dissimilarity penalty calculator for determine the dissimilarity of attributes of alternative unique items from a plurality of selected items.

Claims

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

receiving electronic data indicating a selection of a selected item;

updating, using a computer system, a set of base items to include the selected item, wherein the selected item and the base items each comprise a plurality of attributes;

determining, using the computer system, a set of base attributes from the plurality of attributes of the items in the set of base items to use for determining the similarity of alternative unique items to the set of base items;

selecting, using the computer system, a set of alternative unique items from a plurality of unique items for comparison to the set of base items;

calculating, using the computer system, a first dissimilarity penalty for a first attribute of the set of base attributes for the set of alternative unique items at least partially based on a magnitude of dissimilarity between a value of the first attribute of the set of alternative items and the values of the first attribute of the set of base attributes;

calculating, using the computer system, a second dissimilarity penalty for a second attribute of the set of base attributes for the set of alternative unique items at least partially based on a magnitude of dissimilarity between a value of the second attribute of the set of alternative items and the values of the second attribute of the set of base attributes; and

generating, using the computer system, a recommendation of alternative unique items, the recommendation comprising ranking of at least a portion of the plurality of the set of alternative unique items, the ranking based at least partially on the generated dissimilarity penalties;

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

2. The computer-implemented method of claim 1, wherein the selected item, the base items, and alternative unique items comprise one of the following types of items:

used automobiles, existing homes, real estate, household goods, customized electronics, customized goods.

3. The computer-implemented method of claim 1, wherein the selected item comprises an item in which a user has expressed interest.

4. The computer-implemented method of claim 1, wherein the electronic data indicating a selection of a selected item is received from a user access point system providing a user interface to a user.

5. The computer-implemented method of claim 1, wherein receiving an indication of electronic data indicating a selection of a selected item includes receiving an indication of a level of interest in the selected item.

6. The computer-implemented method of claim 5, wherein updating attributes of a plurality of base items based on the attributes of the selected item comprises determining a new plurality of base items based on the level of interest in the selected item and a level of interest in the plurality of base items.

7. The computer-implemented method of claim 1, wherein calculating the first dissimilarity penalty for the first attribute of the set of base attributes comprises determining a difference between the first attribute of the alternative unique items and a range of values between the maximum and minimum values of the first attribute of the set of base attributes, wherein a minimum penalty is generated if the first attribute of a particular unique item falls within the range of the values.

8. The computer-implemented method of claim 7, wherein the first attribute is the price of the alternative unique items and the plurality of base items.

9. The computer-implemented method of claim 1, wherein generating a recommendation of alternative unique items comprises taking a Mahalanobis distance calculation of the dissimilarity penalties of the attributes of the alternative unique items and the set of base attributes.

10. The computer-implemented method of claim 1, wherein the plurality of alternative unique items comprises at least 100 alternative unique items, and the generating a recommendation of alternative unique items occurs substantially in real time.

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

calculating, using a collaborative filter, a probability score for each of the plurality of alternative unique items, the probability score representing an estimated probability that a user will be interested in the alternative unique item based on the user'"'"'s preferences;

wherein generating the dissimilarity penalty further comprises combining the probability score with at least the first dissimilarity penalty and the second dissimilarity penalty.

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

logging, using the computer system, interactions of users with a plurality of unique items, wherein the logging comprises electronically monitoring actions of the users interacting with one or more item listing systems presenting for sale the plurality of unique items, and wherein the logging produces data configured to be used to determine a level of interest of the users in the selected items.

13. A recommendation system for generating recommendations of alternative unique items, the recommendation system comprising:

an items information database configured to store data relating to unique items;

a penalty computation engine configured to calculate a dissimilarity penalty, the dissimilarity penalty at least partially generated based on a magnitude of dissimilarity between a plurality of selected items and an alternative item;

a recommendation compilation engine configured to generate a recommendation of alternative unique items, wherein the recommendation compilation engine is configured to electronically communicate with the penalty computation engine to calculate dissimilarity penalties for each of a plurality of alternative unique items, the recommendation of alternative unique items comprising a ranking of at least a portion of the plurality of alternative unique items, the ranking based at least partially on the calculated dissimilarity penalties; and

one or more computers configured to operate the recommendation compilation engine, wherein the one or more computers comprises a computer processor and an electronic storage medium.

14. The recommendation system of claim 13, wherein the plurality of selected items and alternative unique items comprise one of the following types of items:

used automobiles, existing homes, real estate, household goods, customized electronics, customized goods.

15. The recommendation system of claim 13, wherein the plurality of selected items comprises items the recommendation system has received an indication that user has selected with an item selection receiver of a user access point system.

16. The recommendation system of claim 13, wherein the recommendation compilation engine is further configured to receive selected item data, the selected item data being related to a plurality of attributes of the selected item, the plurality of attributes comprising at least one customization attribute of the selected item and at least one condition attribute of the selected item.

17. The recommendation system of claim 13, wherein the penalty computation engine calculates the dissimilarity penalty based on attributes of the alternative item compared to attributes of the plurality of selected items.

18. The recommendation system of claim 13, further comprising:

a data collection engine configured to electronically monitor, substantially in real time, user interactions with a plurality of unique items and to generate data based on the user interactions for use by the penalty computation engine in calculating dissimilarity penalties.

19. The recommendation system of claim 18, wherein the data collection engine being configured to electronically monitor, substantially in real time, user interactions, comprises being configured to electronically monitor a plurality of item listing services.

20. A computer-implemented method for generating recommended prices of unique items, the computer-implemented method comprising:

receiving, at a computer system, electronic data indicating attributes for a new listing;

selecting, using the computer system, a set of alternative unique items from a plurality of unique items for comparison to the attributes for the new listing;

calculating, using the computer system, a dissimilarity penalty for each item in the set of alternative unique items, the dissimilarity penalty representing a similarity between the attributes for the new listing and attributes of the alternative unique items;

generating, using the computer system, a set of similar alternative unique items comprising at least a portion of the set of alternative unique items, the set of similar alternative unique items generated based at least partially on the calculated dissimilarity penalties;

generating, using the computer system, a recommended price for the new listing based at least partially on prices associated with the set of similar alternative unique items;

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