Search relevance

Patent metadata

Field	Value
Publication number	US-2016378771-A1
Application number	US-201615260130-A
Country	US
Kind code	A1
Filing date	Sep 8, 2016
Priority date	Apr 30, 2013
Publication date	Dec 29, 2016
Grant date	—

Abstract

The present invention extends to methods, systems, and computer program products for training a classification model to predict categories. In one implementation, a method identifies category mappings generated for dominant queries associated with a query log. The method identifies mappings between a first set of queries and categories shown for the first set of queries, and identifies mappings between a second set of queries and clicked products for the second set of queries. A classification model is trained based on the mappings generated for dominant queries, the mappings between queries and the shown categories, and the mappings between queries and the clicked products.

First claim

What is claimed is:

1 . At a computer system, the computer system including one or more processors and system memory, the computer system communicatively coupled to a query log, the query log including query records for e-commerce queries executed against a product database, each query record containing: one or more categories that were used as search terms, query results from submitting the search terms in a query of the product database, and click through information indicating products that were selected from among the query results, the product database using a plurality of different categories to categorize products, the one or more categories selected from among the plurality of categories, a method for classifying e-commerce queries to generate category mappings, the method comprising:
identifying in the query log query records that are within a specified date range with click through information that indicates that one or more selected products were selected from corresponding query results and; identifying in the query log query records that are within the specified date range with display information that indicates that one or more displayed products were displayed; for each of the one or more categories selected from among the plurality of categories:
calculating a selection rate of each of the one or more selected products selected among at least one corresponding query result returned in response to a query of the category wherein the at least one corresponding query result is from the corresponding query results;
calculating a product display rate for the one or more selected products selected from among the at least one corresponding query result returned in response to the query of the category;
identifying a mapping between the query of the category and the displayed products;
identifying a mapping between the query of the category and the selected products;
calculating a category score of the category based on a first number of times the category was shown and a second number of times the category was clicked;
ranking the one or more categories based on the category score of the category for each of the one or more categories; and use results from the ranking of the one or more categories to respond to at least one query from at least one online consumer to facilitate identification of products of interest to the at least one online consumer.

2 . The method of claim 1 further comprising applying a confidence interval to products in the category to remove bias from the category score calculations.

3 . The method of claim 2 further comprising varying a formula for the confidence interval.

4 . The method of claim 1 wherein the specified date range is six months.

5 . The method of claim 4 further comprising applying equal weighting to all query records, regardless of date, to calculate category scores.

6 . The method of claim 4 further comprising applying increased weighting to recent query records to calculate category scores.

7 . The method of claim 1 , wherein ranking the one or more categories based on calculated category scores comprises assigning each of the one or more categories to a category type based on the calculated category scores.

8 . The method of claim 1 wherein calculating a selection rate for a product comprises calculating a selection rate based on a plurality of categories to which the product is assigned.

9 . The method of claim 8 wherein calculating the selection rate comprises considering the product's assigned primary category.

10 . The method of claim 1 further comprising parsing additional query record details including identifying: products shown to the user, whether or not a product was added to the cart, whether or not the product was ordered, the order number, the product's primary and other category mappings, and position of the product in the search results.

11 . The method of claim 10 further comprising modifying the click through information to consider the additional query record details including add to cart ratio, order ratio, and the position of the product in the search results.

12 . The method of claim 1 further comprising, prior to calculating a selection rate for the one or more products, qualifying the one or more products from among a plurality of products, the plurality of products selected from among at least one corresponding query result returned in response to a query of the category, the one or more products qualified by having: a minimum number of clicks and a minimum number of impressions.

13 . A computer system that classifies e-commerce queries, the computer system comprising:
one or more processors; system memory; and one or more computer storage media having stored thereon computer-executable instructions to:
mine a query log for query records within a specified date range that identify consumer products with shown product information and click through information;
determine a category associated with each of the consumer products;
for each category associated with each of the consumer products:
calculate a display rate and a selection rate for at least one product selected from among at least one corresponding query result returned in response to a query of the category;
calculate a category score based on a first number of times the category is shown and a second number of times the category is clicked; and

rank categories based on the category score for each of the categories to establish a ranking of the categories; and
facilitate online shopping for customers using the ranking of the categories.

14 . The system of claim 13 further comprising computer-executable instructions to apply a confidence interval to remove the bias from the category score.

15 . The system of claim 13 wherein the specified date range is six months.

16 . The system of claim 13 , further comprising computer-executable instructions to calculate and use to establish the ranking: an added to cart ratio, an order ratio, and product position signals.

17 . A computer implemented method comprising:
identifying, using one or more processors, a first set of query records in a consumer product query log with click through information that indicates that a second set of one or more consumer products were selected from consumer product query results; identifying, using the one or more processors, a third set of query records in the consumer product query log with display information that indicates that a fourth set of one or more consumer products were displayed; determining a product category associated with each of a fifth set of one or more consumer products, wherein the fifth set of one or more consumer products includes the second set of one or more consumer products and the fourth set of one or more consumer products; calculating a display rate and a selection rate for each of the second set of one or more consumer products; calculating, using the one or more processors, a category score for the product category associated with each of the fifth set of one or more consumer products, wherein the category score is based on a number of times the product category was shown and a number of times the product category was clicked; and identifying products of interest to online shopping customers using the category score.

18 . The method of claim 17 further comprising assigning a category type to categories associated with each of the products.

19 . The method of claim 17 further comprising: identifying mappings between a six

Assignees

• Wal Mart Stores Inc

Inventors

• Tholiya Namrata Pramodkumar
• Gattani Abhishek

Classifications

• G06F17/30598

  Physics · mapped topic

• G06Q30/0201

  Market modelling; Market analysis; Collecting market data · CPC title

• G06F17/30864

  Physics · mapped topic

• G06F17/30648

  Physics · mapped topic

• G06F17/30696

  Physics · mapped topic