US 12,443,610 B2
Systems and methods for generating trending search queries using predictive search engines
Junchao Zheng, Jersey City, NJ (US); Chen Song, Princeton, NJ (US); Jun Zhao, Jersey City, NJ (US); and Zheng Yan, Short Hills, NJ (US)
Assigned to WALMART APOLLO, LLC, Bentonville, AR (US)
Filed by Walmart Apollo, LLC, Bentonville, AR (US)
Filed on Jan. 30, 2023, as Appl. No. 18/103,264.
Prior Publication US 2024/0256555 A1, Aug. 1, 2024
Int. Cl. G06F 16/24 (2019.01); G06F 16/2457 (2019.01); G06F 16/9535 (2019.01); G06F 16/9538 (2019.01); G06Q 30/0601 (2023.01)
CPC G06F 16/24578 (2019.01) [G06F 16/9535 (2019.01); G06F 16/9538 (2019.01); G06Q 30/0633 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
one or more processors; and
one or more non-transitory computer-readable storage devices storing computing instructions that, when run on the one or more processors, cause the one or more processors to:
receive historical search queries submitted via a search engine on an electronic platform;
generate a first set of user engagement metrics corresponding to a first historical time period for a historical search query of the historical search queries,
wherein the first set of user engagement metrics includes three or more of:
a first typeahead impression metric indicating how many times the historical search query was presented as a typeahead query during the first historical time period,
a first explicit query impression metric indicating how many times the historical search query was entered via a query input field during the first historical time period,
a first click metric indicating how many times an item was selected after the historical search query was submitted during the first historical time period and after the item was presented in search results from the historical search query,
a first add-to-cart metric indicating how many times the item was added to an electronic shopping cart after the historical search query was submitted during the first historical time period and after the item was presented in search results from the historical search query, or
a first order metric indicating how many times the item was ordered after the historical search query was submitted during the first historical time period and after the item was presented in search results from the historical search query;
generate a second set of user engagement metrics corresponding to a second historical time period for the historical search query,
wherein the first historical time period is shorter than the second historical time period, and wherein the second set of user engagement metrics includes three or more of:
a second typeahead impression metric indicating how many times the historical search query was presented as a typeahead query during the second historical time period,
a second explicit query impression metric indicating how many times the historical search query was entered via a query input field during the second historical time period,
a second click metric indicating how many times an item was selected after the historical search query was submitted during the second historical time period and after the item was presented in search results from the historical search query,
a second add-to-cart metric indicating how many times the item was added to an electronic shopping cart after the historical search query was submitted during the second historical time period and after the item was presented in search results from the historical search query, or
a second order metric indicating how many times the item was ordered after the historical search query was submitted during the second historical time period and after the item was presented in search results from the historical search query;
generate significance scores for the historical search query based, at least in part, on the first set of user engagement metrics and the second set of user engagement metrics;
compare each of the significance scores associated with the historical search query to a threshold;
include the historical search query in trending search queries, within the historical search queries, based, at least in part, on each of the significance scores associated with the historical search query satisfying the threshold; and
generate a search engine interface, for the search engine, that displays the trending search queries to a user before the user types a search query into a query input field of the search engine interface.