System and method for mining category aspect information

• US 10,459,919 B2
• Filed: 12/11/2016
• Issued: 10/29/2019
• Est. Priority Date: 06/08/2011
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Zhong, Qian, Song, Guanglei, Chen, Zhaohui
• Primary Examiner(s): Gebresenbet, Dinku W
• Application Number:
  US15/375,144
• Publication Number:
  US 20170091275A1
• Time in Patent Office: 1,052 Days
• CPC Class Codes: G06F 16/2246 Trees, e.g. B+trees
• CPC Class Codes: G06F 16/2282 Tablespace storage structures; Management thereof
• CPC Class Codes: G06F 16/24544 Join order optimisation
• CPC Class Codes: G06F 16/2456 Join operations
• CPC Class Codes: G06F 16/24575 using context
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

In various example embodiments, systems and methods for providing category aspect information by mining historical data is provided. In example embodiments, a table comprising joined data is accessed. The table includes historical data that comprises user behavior data based on actions performed with past queries by users, data describing publication, and a determined category for each publication. Demand scores based on the joined data are determined. The determined demand scores are used to determine most relevant aspect name and aspect value pairs for a category. Publications having at least one of the most relevant aspect name and aspect value pairs are displayed visually distinguished from less relevant publications that exclude the most relevant aspect name and aspect value pairs.

Claims

1. A method comprising:

accessing, by a hardware processor, a table comprising joined data, the joined data including historical data comprising data based on actions performed by users, data describing each publication, and a determined category for each publication;

determining, by an aggregation module, demand scores based on the joined data;

determining, by an aspect determination module, most relevant aspect name/value pairs for a category based on the determined demand scores;

determining a plurality of publications having at least one of the most relevant aspect name/value pairs for the category, a publication of the plurality of publications being created with at least one of the most relevant aspect name/value pairs in response to a message indicating an aspect name of one of the most relevant aspect name/value pairs and suggesting a user input a corresponding aspect value in response to the user creating the publication; and

causing, by a publication engine, the plurality of publications having at least one of the most relevant aspect name/value pairs to be displayed visually distinguished from less relevant publications that exclude the most relevant aspect name/value pairs for the category.

2. The method of claim 1, wherein the causing the plurality of publications to be displayed visually distinguished comprises causing the plurality of publications to be displayed bolded or highlighted.

3. The method of claim 1, further comprising, in response to an indication that a user is creating a new publication within the category, transmitting the message for display on a device of the user, the message indicating the aspect name of one of the most relevant aspect name/value pairs and suggesting the user provide a corresponding aspect value in the new publication.

4. The method of claim 1, further comprising:

compiling the historical data;

accessing the data describing each publication; and

creating the table by joining the historical data with the data describing each publication and the determined category of each listing.

5. The method of claim 2, wherein the determining the most relevant aspect name/value pairs comprises determining an aspect name of one of the most relevant aspect name/value pairs by sorting the determined demand scores in descending order.

6. The method of claim 5, wherein the determining the most relevant aspect name/value pairs comprises sorting the demand scores for the aspect name to determine at least one most relevant aspect value for the aspect name.

7. The method of claim 1, wherein the table comprises categories, aspect names, aspect values, and action counts.

8. The method of claim 1, herein the determining the demand scores is a function of two or more action counts selected from the group consisting of impression counts, click action counts, Buy-it-Now action counts, bid action counts, watch action counts, ask action counts, and offer action counts.

9. A hardware storage device storing instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:

accessing a table comprising joined data, the joined data including historical data comprising data based on actions performed by users, data describing each publication, and a determined category for each publication;

determining demand scores based on the joined data;

determining most relevant aspect/value pairs for a category based on the determined demand scores;

determining a plurality of publications having at least one of the most relevant aspect name/value pairs for the category, a publication of the plurality of publications being created with at least one of the most relevant aspect name/value pairs in response to a message indicating an aspect name of one of the most relevant aspect name/value pairs and suggesting a user input a corresponding aspect value in response to the user creating the publication; and

causing the publications having at least one of the most relevant aspect name/value pairs to be displayed visually distinguished from less relevant publications that exclude the most relevant aspect name/value pairs for the category.

10. The hardware storage device of claim 9, wherein the operations further comprise, in response to an indication that a user is creating a new publication within the category, transmitting a message for display on a device of the user, the message indicating an aspect name of one of the most relevant aspect name/value pairs and suggesting the user provide a corresponding aspect value in the new publication.

11. The hardware storage device of claim 9, wherein the operations further comprise:

compiling the historical data;

accessing the data describing each publication; and

creating the table by joining the historical data with the data describing each publication and the determined category of each listing.

12. The hardware storage device of claim 9, wherein the determining the most relevant aspect name/value pairs comprises determining an aspect name of one of the most relevant aspect name/value pairs by sorting the determined demand scores in descending order.

13. The hardware storage device of claim 12, wherein the determining the most relevant aspect name/value pairs comprises sorting the demand scores for the aspect name to determine at least one most relevant aspect value for the aspect name.

14. The hardware storage device of claim 9, wherein the determining the demand scores is a function of two or more action counts selected from the group consisting of impression counts, click action counts, Buy-It-Now action counts, bid action counts, watch action counts, ask action counts, and offer action counts.

15. A system comprising:

one or more hardware processors; and

a storage device storing instructions that, when executed by the one or more hardware processors, causes the one or more hardware processors to perform operations comprising;

accessing a table comprising joined data, the joined data including historical data comprising data based on actions performed by users, data describing each publication, and a determined category for each publication;

determining demand scores based on the joined data;

determining most relevant aspect name/value pairs for a category based on the determined demand scores;

determining a plurality of publications having at least one of the most relevant aspect name/value pairs for the category, a publication of the plurality of publications being created with at least one of the most relevant aspect name/value pairs in response to a message indicating an aspect name of one of the most relevant aspect name/value pairs and suggesting a user input a corresponding aspect value in response to the user creating the publication; and

causing the publications having at least one of the most relevant aspect name/value pairs to be displayed visually distinguished from less relevant publications that exclude the most relevant aspect name/value pairs for the category.

16. The system of claim 15, wherein the operations further comprise, in response to an indication that a user is creating a new publication within the category, transmitting a message for display on a device of the user, the message indicating an aspect name of one of the most relevant aspect name/value pairs and suggesting the user provide a corresponding aspect value in the new publication.

17. The system of claim 15, wherein the operations further comprise:

compiling the historical data;

accessing the data describing each publication; and

creating the table by joining the historical data with the data describing each publication and the determined category of each listing.

18. The system of claim 15, wherein the determining the most relevant aspect name/value pairs comprises determining an aspect name of one of the most relevant aspect name/value pairs by sorting the determined demand scores in descending order.

19. The system of claim 18, wherein the determining the most relevant aspect name/value pairs comprises sorting the demand scores for the aspect name to determine at least one most relevant aspect value for the aspect name.

20. The system of claim 15, wherein the determining the demand scores is a function of two or more action counts selected from the group consisting of impression counts, click action counts, Buy-It-Now action counts, bid action counts, watch action counts, ask action counts, and offer action counts.