	US 12,093,988 B2
	System, method, and computer readable medium for improving ecommerce search ranking via query-price affinity values
Xinyi Liu, Secaucus, NJ (US); Varun Joshi, Hoboken, NJ (US); Jungwoo Han, Jersey City, NJ (US); Cun Mu, Jersey City, NJ (US); and Rajyashree Mukherjee, San Carlos, CA (US)
Assigned to Walmart Apollo, LLC, Bentonville, AR (US)
Filed by Walmart Apollo, LLC, Bentonville, AR (US)
Filed on May 5, 2021, as Appl. No. 17/308,477.
Prior Publication US 2022/0358559 A1, Nov. 10, 2022
Int. Cl. G06Q 30/06 (2023.01); G06F 16/2457 (2019.01); G06F 16/9535 (2019.01); G06F 16/9538 (2019.01); G06Q 30/0601 (2023.01)

CPC G06Q 30/0631 (2013.01) [G06F 16/24578 (2019.01); G06F 16/9535 (2019.01); G06F 16/9538 (2019.01)]

18 Claims

1. A system comprising:

a query database comprising a lookup table;

a non-transitory medium having instructions stored thereon; and

a processor configured to read the instructions to:

parse a received search query using a machine learning algorithm to identify one or more search terms associated with the search query;

classify the received search query using a query classification model, wherein the query classification model comprises a natural language process and a k-means clustering process, and wherein the query classification model is trained by an iterative training process based on a training dataset including historical queries to classify the received search query into one of the plurality of existing query categories or a new query category;

obtain a plurality of items based on the one or more search terms associated with the received search query and a corresponding value of each item in the plurality of items;

in response to classifying the received search query into one of the plurality of existing query categories:

retrieve a value distribution for the received search query, wherein the value distribution for the received search query is based on aggregated interaction data for each query in the query category of the received search query;

for each item of the plurality of items, generate a feature value based on the value distribution and the corresponding value of the item;

generate a ranked set of items by ranking the plurality of items by the feature value;

transmit instructions to a customer device that causes the customer device to display the ranked set of items in a user interface in a ranked arrangement based on the corresponding feature value, wherein the instructions include the ranked set of items;

store the received search query in the query database; and

update the lookup table to associate the received search query with the one of the plurality of existing query categories;

in response to classifying the received search query into the new query category:

transmit instructions to the customer device that causes the customer device to display the plurality of items;

store the received search query in the query database; and

update the lookup table to include the new query category; and

generate an update value distribution for an updated query category in the query database based on updated aggregated interaction data including interaction data associated with the received search query.