Identifying product groups in ecommerce
View Patent ↗Systems and methods are disclosed herein for supplementing product records with product groups that are relevant to the product records. Queries form users may be analyzed to extract keywords. Search results for keywords are evaluated to determine category consistency among product records, including such values as entropy and taxonomy depth. Those keywords with search results having adequate category consistency are selected as product groups and the search results associated with the product groups. Product groups are associated with product records according to a random walk of a graph having as nodes products and product groups and links representing belonging of a product to a product group. Product groups may be selected based on a transition probability based on a random walk and a quality score based on usage of a product group page for the product group.
1. A method for supplementing product information, the method comprising, by a computer system:
receiving a plurality of keywords;
for each keyword of at least a portion of the plurality of keywords, obtaining a result set for the each keyword by searching a product taxonomy;
obtaining a filtered set of keywords from the at least the portion of the plurality of keywords by filtering a set of keywords according to categories of result sets for the set of keywords, wherein the obtaining comprises, for the each keyword of the at least the portion of the plurality of keywords and the result set for the each keyword:
evaluating a categorical similarity among products of the result set for the each keyword, wherein the evaluating comprises calculating a category consistency of the result set for the each keyword, and wherein the calculating comprises calculating the category consistency as n/(N-n), where n is a number of products in the result set for the each keyword and belonging to a most common category in categories represented in the result set for the each keyword and where N is a total number of products in the result set for the each keyword;
assigning a score to the each keyword according to the categorical similarity; and
filtering the at least the portion of the plurality of keywords according to scores assigned thereto to obtain the filtered set of keywords; and
defining product groups for at least a portion of the filtered set of keywords, each product group having associated therewith a keyword from the filtered set of keywords and one or more products from the result set for the keyword from the filtered set of keywords.
2. The method of claim 1 , wherein the plurality of keywords are extracted from user queries.
3. The method of claim 1 , wherein receiving the plurality of keywords comprises:
receiving a plurality of queries;
de-duplicating constituent terms of the plurality of queries and transforming the constituent terms into canonical forms to obtain a processed set of terms; and
using the processed set of terms as the plurality of keywords.
4. The method of claim 1 , wherein evaluating categorical similarity among products of the result set for the each keyword further comprises determining a distance in the product taxonomy from a root node in the product taxonomy of a most common category of the product taxonomy represented in the result set for the each keyword.
5. The method of claim 4 , wherein evaluating categorical similarity among products of the result set for the each keyword further comprises:
determining a category consistency of the result set for the each keyword; and
wherein assigning a score to the each keyword comprises assigning a score corresponding to a combination of the distance and the category consistency.
6. The method of claim 1 , further comprising for each keyword of the filtered set of keywords and each result set for the keyword of the result sets corresponding to the each keyword, removing from the result sets products that do not belong to a most common category represented in the result set for the keyword.
7. The method of claim 1 , further comprising associating one or more product groups of the defined product groups with products of the product taxonomy.
8. A system for supplementing product information, the system comprising one or more processors and one or more memory devices operably coupled to the one or more processors, the one or more memory devices storing executable and operational data effective to cause the one or more processors to:
receive a plurality of keywords;
for each keyword of at least a portion of the plurality of keywords, obtain a result set for the each keyword by searching a product taxonomy;
obtain a filtered set of keywords from the at least the portion of the plurality of keywords by filtering a set of keywords according to categories of result sets for the set of keywords, wherein the executable and operational data are further effective to cause the one or more processors to obtain the filtered set of keywords from the at least the portion of the plurality of keywords by filtering the set of keywords according to categories of the result sets for the set of keywords by, for each keyword of the at least the portion of the plurality of keywords and the result set for the each keyword performing:
evaluating categorical similarity among products of the result set for the each keyword, wherein the executable and operational data are further effective to cause the one or more processors to evaluate the categorical similarity among products of the result set for the each keyword by calculating category consistency of the result set for the each keyword and wherein the executable and operational data are further effective to cause the one or more processors to calculate the category consistency of the result set for the each keyword by calculating the category consistency as n/(N−n), where n is a number of products in the result set for the each keyword and belonging to a most common category in categories represented in the result set for the each keyword and where N is a total number of products in the result set for the each keyword;
assigning a score to the each keyword according to the categorical similarity; and
filtering the at least the portion of the plurality of keywords according to the scores assigned thereto to obtain the filtered set of keywords; and
define product groups for at least a portion of the filtered set of keywords, each product group having associated therewith a keyword from the filtered set of keywords and one or more products from the result set for the each keyword from the filtered set of keywords.
9. The system of claim 8 , wherein the plurality of keywords are extracted from user queries.
10. The system of claim 8 , wherein the executable and operational data are further effective to cause the one or more processors to receive the plurality of keywords by:
receiving a plurality of queries;
de-duplicating constituent terms of the plurality of queries and transforming the constituent terms into canonical forms to obtain a processed set of terms; and
using the processed set of terms as the plurality of keywords.
11. The system of claim 8 , wherein the executable and operational data are further effective to cause the one or more processors to evaluate categorical similarity among products of the result set for the each keyword by determining a distance in the product taxonomy from a root node in the product taxonomy of a most common category of the product taxonomy represented in the result set for the each keyword.
12. The system of claim 11 , wherein the executable and operational data are further effective to cause the one or more processors to evaluate categorical similarity among products of the result set for the each keyword by:
determining a category consistency of the result set for the each keyword; and
wherein assigning a score to the each keyword comprises assigning a score corresponding to a combination of the distance and the category consistency.
13. The system of claim 8 , wherein the executable and operational data are further effective to cause the one or more processors to, for each keyword of the filtered set of keywords and each result set of the each keyword of the result sets corresponding to the each keyword, remove from the result sets products that do not belong to a most common category represented in the result set of the each keyword.
14. The system of claim 8 , wherein the executable and operational data are further effective to cause the one or more processors to associate one or more product groups of the defined product groups with products of the product taxonomy.