SYSTEMS, APPARATUSES, AND METHODS FOR PROVIDING A RANKING BASED RECOMMENDATION

• US 20160063065A1
• Filed: 12/30/2014
• Published: 03/03/2016
• Est. Priority Date: 08/28/2014
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Khatri, Chandra, Luan, Steven Hui, Tanaka, Michael, Boinapalli, Praveen K.
• Granted Patent: US 11,809,501 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

One or more of the systems, apparatuses, or methods discussed herein can include a quality score for a plurality of item listings or collections of item listings. Data sparseness can be avoided, as the quality score is based on inherent properties of the listing. An item listing can be recommended to a user based on the quality score. In one or more embodiments, a method can include determining a plurality of quality scores including a quality score for each of a plurality of item listings or a plurality of collections of item listings, the quality scores determined independent of a user'"'"'s attributes and independent of the user'"'"'s contextual information, the contextual information corresponding to details of the user'"'"'s access to a website, and recommending an item listing or collection of item listings to a user based on the quality scores and the contextual information.

Claims

1. A system comprising:

a contextual information module configured to determine contextual information, the contextual information including parameters associated with a device used in accessing a website hosting a plurality of item listings;

a quality score module executable by one or more processors and configured to calculate a quality score for each item listing of the plurality of item listings, the quality score independent of attributes of a user and independent of the contextual information, the quality score indicating a probability of a user engagement with a respective item listing;

a ranking module configured to receive the contextual information and the quality scores and rank each of the plurality of item listings based on the contextual information and the quality scores; and

a recommendation module configured to automatically select one or more item listings of the plurality of item listings based on the ranking of the item listings and provide information indicative of the selected one or more item listings.

2. The system of claim 1, wherein the quality score module is further configured to determine the quality score using three or more of item listing freshness, item listing quality, item listing image quality, cover image quality of the item listing, top five mean item listing quality listings, top five mean image quality of the item listing, number of item listings in the item listing, age of the item listing in the item listing, item listing last update time, follow counts of the item listing, unfollow counts of the item listing, share counts of the item listing, number of item listing views, and user engagement with the item listing.

3. The system of claim 2, wherein the quality score module is further configured to utilize a random forest classifier technique in determining the quality scores.

4. The system of claim 1, wherein the recommendation module is configured to compare the determined quality scores to a threshold and wherein the item listing associated with the quality score is only selected if the quality score is greater than the threshold.

5. The system of claim 1, wherein the contextual information includes a type of device the user is using to access the website including whether the device is a mobile device.

6. The system of claim 5, wherein:

the ranking module is further configured to pre-filter the item listings based on the contextual information offline;

the ranking module is further configured to determine whether user attributes are available for the user and in response to determining user attributes are available, further filter the pre-filtered item listings based on the user attributes and filter the further filtered item listings based on the determined quality scores to create a first filtered set of item listings and wherein the recommendation module is further configured to select an item listing from the first filtered set; and

the ranking module is further configured to determine whether user attributes are available for the user and in response to determining the user attributes are not available, further filter the pre-filtered item listings based on the quality scores to create a second filtered set and wherein the recommendation module is further configured to select an item listing from the second filtered set.

7. The system of claim 5, wherein:

the ranking module is further configured to determine if user attributes are available and in response to determining user attributes are not available, the recommendation module is to select an item listing remaining after filtering the plurality of item listings based on the quality score and then filtering based on the contextual information; and

the ranking module is further configured to determine if user attributes are available and in response to determining the user attributes are available, to recommend an item listing remaining after filtering the plurality of item listings based on the determined quality scores and the user attributes, and then filtering based on the contextual information.

8. The system of claim 5, wherein:

the ranking module is further configured to determine if user attributes are available and in response to determining user attributes are not available, to select an item listing remaining after filtering the plurality of item listings based on the contextual information then filtering based on the quality score; and

the ranking module is further configured to determine if user attributes are available and in response to determining the user attributes are available, to select an item listing remaining after filtering the plurality of item listings based on user attributes and contextual information and then filtering based on the determined quality scores.

9. The system of claim 1, wherein the ranking module is further configured to determine the quality scores using a logistic regression technique.

10. The system of claim 1, wherein the ranking module is further configured to determine the quality scores using a stochastic gradient descent technique.

11. The system of claim 1, wherein the ranking module is further configured to determine the quality scores using a Gaussian naï

ve Byes technique.

12. The system of claim 1, wherein the ranking module is further configured to determine the quality scores using an Adaboost technique.

13. A method comprising:

determining, using a computer processor, contextual information, the contextual information including parameters associated with a device used in accessing a website hosting a plurality of item listings;

determining a quality score for each of a plurality of item listings, the quality score independent of a user'"'"'s attributes and independent of the contextual information of the user, the quality score indicating a probability of a user engagement with a respective item listing;

ranking each of the plurality of item listings based on the contextual information and the quality scores; and

automatically selecting one or more item listings of the plurality of item listing based on the ranking of the item listings; and

providing information indicative of the selected item listing.

14. The method of claim 13, wherein determining the quality scores includes determining the quality scores using a random forest classifier technique, a logistic regression technique, a stochastic gradient descent technique, Gaussian naï

ve Byes technique, or an Adaboost technique.

15. The method of claim 14, wherein ranking each of the plurality of item listings includes:

pre-filtering the item listings based on the contextual information offline;

determining whether user attributes are available for the user;

in response to determining user attributes are available, further filtering the pre-filtered item listings based on the user attributes and filtering the further filtered item listings based on the determined quality scores to create a first filtered set and selecting an item listing includes selecting an item listing from the first filtered set; and

in response to determining the user attributes are not available, further filtering the pre-filtered item listings based on the quality scores to create a second filtered set and wherein selecting an item listing includes selecting an item listing from the second filtered set.

16. The method of claim 14, wherein ranking each of the plurality of item listings includes:

determining if user attributes are available;

in response to determining user attributes are not available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the quality score and then filtering based on the contextual information; and

in response to determining the user attributes are available, selecting the item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the determined quality scores and the user attributes and then filtering based on the contextual information.

17. The method of claim 14, wherein ranking each of the plurality of item listings includes:

determining if user attributes are available;

in response to determining user attributes are not available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the contextual information then filtering based on the quality score; and

in response to determining the user attributes are available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on user attributes and contextual information and then filtering based on the determined quality scores.

18. A non-transitory machine-readable storage medium embodying instructions which, when executed by a machine, cause the machine to execute operations comprising:

determining contextual information, the contextual information including parameters associated with a device used in accessing a website hosting a plurality of item listings;

determining a quality score for each of a plurality of item listings using a random forest classifier technique, the quality score independent of a user'"'"'s attributes and independent of the contextual information of the user, the quality score indicating a probability of a user engagement with a respective item listing;

ranking each of the plurality of item listings based on the contextual information and the quality scores; and

automatically selecting one or more item listings of the plurality of item listing based on the ranking of the item listings; and

providing information indicative of the selected item listing.

19. The non-transitory machine readable storage medium of claim 18, wherein the instructions for ranking each of the plurality of item listings include instructions, which when executed by the machine, cause the machine to perform operations comprising:

determining if user attributes are available;

in response to determining user attributes are not available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the quality score and then filtering based on the contextual information; and

in response to determining the user attributes are available, selecting the item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the determined quality scores and the user attributes and then filtering based on the contextual information.

20. The non-transitory machine readable storage medium of claim 18, wherein the instructions for ranking each of the plurality of item listings include instructions, which when executed by the machine, cause the machine to perform operations comprising:

determining if user attributes are available;

in response to determining user attributes are not available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on the contextual information then filtering based on the quality score; and

in response to determining the user attributes are available, selecting an item listing includes selecting an item listing remaining after filtering the plurality of item listings based on user attributes and contextual information and then filtering based on the determined quality scores.