US-12620016-B2 - Electronic device and method for recommending item to user

Abstract

An electronic device is provided. The electronic device includes a communication circuit for communicating with a server, a display, at least one processor connected, and a memory, wherein the memory stores instructions causing, when executed, the at least one processor to collect an intermediate operation value corresponding to an item to which a user responds among first items introduced to the user through the display, store the collected intermediate operation value, receive item information about second items and intermediate operation values, input, as an input value, the intermediate operation values received from the server, input first user information indicating a profile of the user and the user response item vector, obtain scores assigned to each of the second items from a result value output from the recommendation model, and provide the item information to the user through the display on the basis of the scores.

Inventors

• Hwangki MIN
• Yoonyoung NAM
• Jeongsoo LEE
• Gunbong LEE
• Changku LEE

Assignees

• SAMSUNG ELECTRONICS CO., LTD.

Dates

Publication Date

20260505

Application Date

20231108

Priority Date

20210811

Claims (19)

1. 1 . An electronic device comprising: a communication circuit configured to communicate with a server; a display; at least one processor connected to the communication circuit and the display; and memory connected to the at least one processor, wherein the memory stores instructions that, when executed by the at least one processor, cause the electronic device to: collect an intermediate operation value corresponding to an item to which a user has responded among first items introduced to the user through the display, and store, in the memory, the collected intermediate operation value as a user response item vector, receive item information about second items to be introduced to the user and intermediate operation values corresponding to the respective second items from the server through the communication circuit, the second items identified by the server using an estimated profile of the user, input, as an input value, the intermediate operation values received from the server to a recommendation model trained using an artificial intelligence algorithm, acquiring a profile of the user containing personal information of the user stored in the memory without exposing the personal information outside of the electronic device, input, to the recommendation model, first user information indicating the acquired profile of the user and the user response item vector as a parameter for assigning a score to each of the second items, obtain scores assigned to each of the second items from a result value output from the recommendation model by inputting the input value and the parameter to the recommendation model, and provide the item information to the user through the display based on the scores assigned to each of the second items.
2. 2 . The electronic device of claim 1 , wherein the instructions are further configured to cause the electronic device to: receive a user vector corresponding to second user information of the user and represented by a plurality of real values from the server through the communication circuit, and input the user vector as an additional parameter to the recommendation model.
3. 3 . The electronic device of claim 1 , wherein the instructions are further configured to cause the electronic device to: arrange identifiers representing each of the second items in descending order of scores assigned thereto, and display the arranged identifiers on the display.
4. 4 . The electronic device of claim 1 , wherein the instructions are further configured to cause the electronic device to: select items having a score equal to or higher than a designated value from among the second items, and display identifiers representing the selected items on the display.
5. 5 . The electronic device of claim 1 , wherein the instructions are further configured to cause the electronic device to: sequentially input the intermediate operation values received from the server to the recommendation model, and sequentially acquire scores assigned to the second items from the recommendation model.
6. 6 . The electronic device of claim 1 , wherein the recommendation model is configured to: encode the first user information into a user vector represented by a plurality of real values, and assign a score to an item corresponding to the intermediate operation value received from the server through the communication circuit, by using a first user vector.
7. 7 . The electronic device of claim 1 , wherein the recommendation model is configured to: encode the first user information into a first user vector represented by a plurality of real values, receive a second user vector corresponding to second user information of the user and represented by a plurality of real values from the server through the communication circuit, combine the first user vector and the second user vector so as to obtain an end user vector, and assign a score to an item corresponding to the intermediate operation value received from the server through the communication circuit, by using the end user vector.
8. 8 . The electronic device of claim 7 , wherein the recommendation model is configured to use the intermediate operation values received from the server as a parameter for obtaining the end user vector.
9. 9 . The electronic device of claim 1 , wherein the personal information includes at least one of age, gender, address, hobbies, interests, or favorite items.
10. 10 . A recommendation system comprising: a server; and a user device, wherein the server is configured to: identify a recommended item using an estimated profile of a user of the user device without receiving personal information of the user from the user device, encode the recommended item to be introduced to the user of the user device into an intermediate operation value represented by a plurality of real values, and transmit information regarding the recommended item and the intermediate operation value to the user device, and wherein the user device is configured to: encode first user information representing an acquired profile of the user into a first user vector represented by a plurality of real values, the acquired profile of the user containing personal information of the user stored in memory of the user device without exposing the personal information outside of the user device, assign a score to the recommended item by using the first user vector and the intermediate operation value received from the server, and control display of information regarding the recommended item based on the score.
11. 11 . The electronic device of claim 1 , wherein the electronic device comprises a user device and the recommendation model includes an artificial intelligence model that is machine-trained using an artificial intelligence algorithm stored in the memory of the electronic device.
12. 12 . The electronic device of claim 1 , wherein the recommendation model is further configured to assign scores to recommended news items, and wherein the electronic device further comprises a display control model, the display control model configured to: determine a rank of the recommended news items based on the assigned scores, and display titles of the recommended news items according to the rank.
13. 13 . The recommendation system of claim 10 , wherein the estimated profile is estimated based on information about an item to which the user has responded, among previously recommended items, from the user device.
14. 14 . The recommendation system of claim 10 , wherein the server is further configured to: encode second user information of the user into a second user vector represented by a plurality of real values, and transmit the second user vector to the user device.
15. 15 . The recommendation system of claim 14 , wherein the server is further configured to use the intermediate operation value to be transmitted to the user device as a parameter for obtaining the second user vector.
16. 16 . The recommendation system of claim 14 , wherein the user device is further configured to: combine the first user vector with the second user vector received from the server so as to obtain an end user vector, and assign a score to the recommended item by using the end user vector and the intermediate operation value received from the server.
17. 17 . The recommendation system of claim 16 , wherein the user device is further configured to use the intermediate operation value received from the server as a parameter for obtaining the end user vector.
18. 18 . A method performed by an electronic device, the method comprising: collecting an intermediate operation value corresponding to an item to which a user responds among first items introduced to a user through a display of the electronic device, and storing, in a memory of the electronic device, the collected intermediate operation value as a user response item vector; receiving item information about second items to be introduced to the user and intermediate operation values corresponding to the respective second items from a server through a communication circuit of the electronic device, the second items identified by the server using an estimated profile of the user; acquiring a profile of the user containing personal information of the user stored in memory of the electronic device without exposing the personal information outside of the electronic device; inputting, as an input value, the intermediate operation values received from the server to a recommendation model trained using an artificial intelligence algorithm, and inputting, to the recommendation model, first user information indicating the acquired profile of the user and the user response item vector as a parameter for assigning a score to each of the second items; obtaining scores assigned to each of the second items from a result value output from the recommendation model by inputting the input value and the parameter to the recommendation model; and providing the item information to the user through the display based on the scores assigned to each of the second items.
19. 19 . The method of claim 18 , further comprising: receiving a user vector corresponding to second user information of the user and represented by a plurality of real values from the server through the communication circuit; and inputting the user vector as an additional parameter to the recommendation model.

Description

CROSS-REFERENCE TO RELATED APPLICATION(S) This application is a continuation application, claiming priority under § 365(c), of an International application No. PCT/KR2022/003901, filed on Mar. 21, 2022, which is based on and claims the benefit of a Korean patent application number 10-2021-0105872, filed on Aug. 11, 2021, in the Korean Intellectual Property Office, the disclosure of which is incorporated by reference herein in its entirety. BACKGROUND 1. Field The disclosure relates to an electronic device that recommends an item (or content) by using artificial intelligence. 2. Description of Related Art A recommendation system may collect user information representing a user's individual profile and a user's response to an item, and may introduce an item that the user has not encountered yet to the user based on the collected user information. The recommendation system may input, as input values, user information and information about items to an artificial intelligence model with machine learning, and based on a result value (e.g., a score indicating the user's preference for each of the items) output from the artificial intelligence model, the system may arrange the items in descending order and preferentially provide an item with a higher ranking through a display of a user device. The above information is presented as background information only to assist with an understanding of the disclosure. No determination has been made, and no assertion is made, as to whether any of the above might be applicable as prior art with regard to the disclosure. SUMMARY A recommendation system may be constructed in a server or a user device. In case of being constructed in the server, there is a risk that user information (personal profile) may be exposed to the outside. As the number of users to receive recommendation service increases and the recommendation service is required in real time, the cost of constructing the recommendation system in the server may increase. In case of being constructed in the user device, the performance of the recommendation system may be limited due to resources provided in the user device, and the user may have resistance to the resources of the user device (e.g., a processor, a memory) being put to the recommendation system. According to various embodiments, some functions related to user personal information may be decomposed from a recommendation system (e.g., a recommendation model) and performed in an electronic device (e.g., a user device). The electronic device may prevent user information from being exposed to the outside as much as possible and provide a recommendation service at low cost/high efficiency. Aspects of the disclosure are to address at least the above-mentioned problems and/or disadvantages and to provide at least the advantages described below. Accordingly, an aspect of the disclosure is to provide an electronic device that recommends an item (or content) by using artificial intelligence. Additional aspects will be set forth in part in the description which follows and, in part, will be apparent from the description, or may be learned by practice of the presented embodiments. In accordance with an aspect of the disclosure, an electronic device is provided. The electronic device includes a communication circuit configured to communicate with a server, a display, at least one processor connected to the communication circuit and the display, and a memory connected to the at least one processor, wherein the memory is configured to store instructions which, when executed by the at least one processor, cause the at least one processor to collect an intermediate operation value corresponding to an item to which a user has responded among first items introduced to the user through the display and store in the memory, the collected intermediate operation value as a user response item vector, receive item information about second items to be introduced to the user and intermediate operation values corresponding to the respective second items from the server through the communication circuit, input, as an input value, the intermediate operation values received from the server to a recommendation model trained using an artificial intelligence algorithm, input, to the recommendation model, first user information indicating a profile of the user and the user response item vector as a parameter for assigning a score to each of the second items, obtain scores assigned to each of the second items from a result value output from the recommendation model by inputting the input value and the parameter to the recommendation model, and provide the item information to the user through the display based on the scores assigned to each of the second items. In accordance with another aspect of the disclosure, a recommendation system is provided. The recommendation system includes a server, and a user device. The server may be configured to encode a recommended item to be introduced to a use