US 11,954,137 B2
Data generation device and data generation method
Kosei Fume, Kawasaki (JP)
Assigned to KABUSHIKI KAISHA TOSHIBA, Minato-ku (JP)
Filed by KABUSHIKI KAISHA TOSHIBA, Minato-ku (JP)
Filed on Aug. 26, 2020, as Appl. No. 17/003,328.
Claims priority of application No. 2019-222380 (JP), filed on Dec. 9, 2019.
Prior Publication US 2021/0173857 A1, Jun. 10, 2021
Int. Cl. G06F 17/00 (2019.01); G06F 7/00 (2006.01); G06F 16/2457 (2019.01); G06F 16/332 (2019.01); G06F 16/335 (2019.01); G06F 18/2113 (2023.01); G06F 18/2323 (2023.01)
CPC G06F 16/337 (2019.01) [G06F 16/24578 (2019.01); G06F 16/3326 (2019.01); G06F 16/335 (2019.01); G06F 18/2113 (2023.01); G06F 18/2323 (2023.01)] 12 Claims
OG exemplary drawing
 
1. A data generation device comprising:
processing circuitry configured to:
classify each of a plurality of users having attributes similar to one another into one of a plurality of clusters based on user information representing the respective attributes of the users;
perform a first search processing using a first search query that includes one or more keywords extracted from the user information to obtain one or more first search results;
estimate whether the users select each of the first search results based on a similarity between each of the first search results and information representing attributes of the clusters, and generating estimation information representing whether the users select each of the first search results;
generate learning data of a ranking model that determines an order of a plurality of second search results, the learning data including the first search results and the estimation information;
train the ranking model using the learning data;
perform a third search processing using a third search query that includes one or more received keywords to obtain a plurality of third search results;
determine the order of the third search results using the ranking model that is trained using the learning data; and
output the third search results in the order that is determined using the ranking model.