US 11,720,927 B2
Method and apparatus for generating user-ad matching list for online advertisement
Junbeom Kim, Suwon-si (KR); Kimin Oh, Suwon-si (KR); Taeho Hwang, Suwon-si (KR); and Kiwon Choi, Suwon-si (KR)
Assigned to Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed by Samsung Electronics Co., Ltd., Suwon-si (KR)
Filed on Nov. 15, 2021, as Appl. No. 17/526,421.
Application 17/526,421 is a continuation of application No. PCT/KR2021/013935, filed on Oct. 8, 2021.
Claims priority of application No. 10-2021-0004930 (KR), filed on Jan. 13, 2021.
Prior Publication US 2022/0222712 A1, Jul. 14, 2022
Int. Cl. G06Q 30/02 (2023.01); G06Q 30/0251 (2023.01); G06Q 30/0241 (2023.01); G06N 3/045 (2023.01)
CPC G06Q 30/0271 (2013.01) [G06N 3/045 (2023.01); G06Q 30/0255 (2013.01); G06Q 30/0277 (2013.01)] 18 Claims
OG exemplary drawing
 
1. A method, performed by a server, of generating a user-advertisement (user-ad) matching list for online advertisement, the method comprising:
receiving, from an ad impression server, profile information of users and advertisement impression history information of the users, wherein, the users are people who have been exposed to a plurality of advertisements through the ad impression server;
receiving, from a server of a first advertiser, advertisement information regarding an advertisement product and advertisement strategy information including a plurality of conditions for exposing the advertisement;
identifying candidate users for the advertisement of the first advertiser among the users, based on the profile information of the users and the advertisement strategy information;
obtaining a predicted access time when the candidate users will access the ad impression server for each time slot having a certain time interval, by inputting the profile information of the candidate users and the advertisement impression history information to a first artificial intelligence model;
selecting target users to be provided with the advertisement of the first advertiser among the candidate users, based on the predicted access time;
obtaining a preference of each target user for the advertisement product of the first advertiser by inputting the profile information of the target users, the advertisement impression history information, and the advertisement information to a second artificial intelligence model;
determining, based on the preference and advertisement budget of the first advertiser, first target users having a high priority and preliminary target users not included in the first target users as second target users;
matching, based on the preference, the advertisement of the first advertiser with the first target users and the second target users;
generating the user-ad matching list matched by a plurality of target users and a plurality of advertisers by a time period, wherein, the user-ad matching list includes the first advertiser and the first target users for the first advertiser and the second target user for the first advertiser;
receiving, from the ad impression server, an advertisement request after the user-ad matching list is generated, wherein, the advertisement request includes profile information of a first user who accessed the ad impression server; and
transmitting, to the ad impression server, advertisement information in case of the first user is the first target and providing advertisement information after a predetermined time in case of the first user is the second target user.