Granted Patent B2

US 12,547,903 · App. 17/439,758 · Granted Feb 10, 2026

Inventors: Mingyuan Ma (Beijing, CN); Tianxin Liang (Beijing, CN); Wenchu Dong (Beijing, CN); Ken Wen (Beijing, CN)

Assignee: BOE TECHNOLOGY GROUP CO., LTD.

G06N5/022G06F16/21G06F40/295G06N5/04

Abstract

Claims (82)

1 . A human-computer interaction method, comprising:

receiving, by a terminal device, an input question;

extracting, by a terminal device, entities and relationships involved in the input question;

querying, by a server, an answer to the input question in a knowledge graph according to the entities and the relationships;

displaying, by a terminal device, a sub-graph of the knowledge graph, wherein the sub-graph of the knowledge graph comprises entities involved in the input question and the answer, and a directional identifier, wherein the directional identifier is configured to identify a reasoning path corresponding to the query process; and

in an instance in which the sub-graph of the knowledge graph is displayed by the terminal device:

detecting a degree of association between a current input question and a previous dialog entity, and determining whether to delete a previous sub-graph of the knowledge graph based on the degree of association;

in response to determining that the current input question and a previous dialog entity are not relevant, deleting the previous sub-graph of the knowledge graph, obtaining and displaying a new sub-graph of the knowledge graph corresponding to the current input question; and

in response to determining that the current input question and a previous dialog entity are relevant, continuing to add entities and relationships to the previous sub-graph of the knowledge graph,

wherein, in the instance in which the sub-graph of the knowledge graph is displayed by the terminal device, the method further comprises:

acquiring at least one of user attribute data, user behavior data, and operation data, wherein the user attribute data comprises at least one of age, gender, and purchasing power of a user, and the user behavior data comprises historical search data;

acquiring a recommended object according to at least one of the user attribute data, the user behavior data, and the operation data; and

displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph,

wherein displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph comprises:

popping up a message prompt window in response to an operation of a user of clicking the entity, and displaying the recommended object in the message prompt window.

2 . The human-computer interaction method according to claim 1 , wherein the directional identifier is configured to connect entities that are successively analyzed in the query process.

3 . The human-computer interaction method according to claim 2 , wherein:

the input question is a multi-order reasoning question;

in response to the input question being the multi-order reasoning question, the directional identifier comprises a multi-level directional identifier to distinguish the successively corresponding reasoning path of each order in the query process; and

any two of the multi-level directional identifiers have at least one of different colors, sizes, and shapes.

4 . The human-computer interaction method according to claim 1 , wherein the directional identifier is a directional arrow.

5 . The human-computer interaction method according to claim 1 , wherein the directional identifier is represented by a plurality of entities with different attributes that are successively analyzed in the query process, and the attribute comprises at least one of color, size, and shape.

6 . The human-computer interaction method according to claim 1 , wherein, in an instance in which the input question is greater than or equal to a second round of question and answer interaction, displaying the sub-graph of the knowledge graph comprises:

judging whether the input question and the answer involve a new entity compared with the input question and answer of a previous round of interaction;

updating the new entity and the corresponding directional identifier to the sub-graph of the knowledge graph obtained in the previous round of interaction, and displaying an updated sub-graph of the knowledge graph, if it involves a new entity; and

serving the sub-graph of the knowledge graph obtained in the previous round of interaction as the sub-graph of the knowledge graph, and updating the directional identifier, if it does not involve a new entity.

7 . The human-computer interaction method according to claim 6 , further comprising: displaying differentially the reasoning paths corresponding to different input questions based on the directional identifier.

8 . The human-computer interaction method according to claim 1 , wherein displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph comprises:

displaying differentially a target entity by color filling or a symbol mark in the sub-graph of the knowledge graph, wherein the target entity is an entity connected to the recommended object through a relationship; and

receiving a control operation acting on the target entity, and displaying the recommended object based on the control operation.

9 . The human-computer interaction method according to claim 1 , wherein displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph comprises:

displaying the recommended object and a relationship connected to the recommended object with a dashed line, wherein the recommended object is displayed in the sub-graph of the knowledge graph in a form of an entity.

10 . The human-computer interaction method according to claim 1 , wherein displaying a sub-graph of the knowledge graph comprises:

selecting among the entities involved in the input question and the answer according to a preset screening rule; and

displaying the sub-graph of the knowledge graph, wherein the sub-graph of the knowledge graph comprises the selected entity and the directional identifier.

11 . The human-computer interaction method according to claim 10 , wherein selecting among the entities involved in the input question and the answer according to a preset screening rule comprises:

in response to a control operation of a user, selecting a display state of the entity and its related entities according to the control operation,

wherein selecting a display state of the entity and its related entities according to the control operation comprises:

in response to a selection operation acting on an entity, hiding a branch centered on the entity and displaying a hidden icon, wherein the branch comprises all entities, relationships, and directional identifiers connected to the entity; or,

in response to a selection operation acting on a relationship, hiding all the entities connected to the relationship and displaying the hidden icon.

12 . A non-transitory computer-readable storage medium with program instructions stored thereon that, when executed by at least one hardware processor, directs the at least one hardware processor to perform a method for human-computer interaction, comprising:

receiving, by a terminal device, an input question;

extracting, by a terminal device, entities and relationships involved in the input question;

querying, by a server, an answer to the input question in a knowledge graph according to the entities and the relationships;

displaying, by a terminal device, a sub-graph of the knowledge graph, wherein the sub-graph of the knowledge graph comprises entities involved in the input question and the answer, and a directional identifier, wherein the directional identifier is configured to identify a reasoning path corresponding to the query process; and

in an instance in which the sub-graph of the knowledge graph is displayed by the terminal device:

detecting a degree of association between a current input question and a previous dialog entity, and determining whether to delete a previous sub-graph of the knowledge graph based on the degree of association;

in response to determining that the current input question and a previous dialog entity are not relevant, deleting the previous sub-graph of the knowledge graph, obtaining and displaying a new sub-graph of the knowledge graph corresponding to the current input question; and

in response to determining that the current input question and a previous dialog entity are relevant, continuing to add entities and relationships to the previous sub-graph of the knowledge graph,

wherein, in the instance in which the sub-graph of the knowledge graph is displayed by the terminal device, the method further comprises:

acquiring at least one of user attribute data, user behavior data, and operation data, wherein the user attribute data comprises at least one of age, gender, and purchasing power of a user, and the user behavior data comprises historical search data;

acquiring a recommended object according to at least one of the user attribute data, the user behavior data, and the operation data; and

displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph,

wherein displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph comprises:

popping up a message prompt window in response to an operation of a user of clicking the entity, and displaying the recommended object in the message prompt window.

13 . The non-transitory computer-readable storage medium according to claim 12 , wherein the directional identifier is configured to connect entities that are successively analyzed in the query process.

14 . The non-transitory computer-readable storage medium according to claim 13 , wherein:

in response to the input question being a multi-order reasoning question, the directional identifier comprises a multi-level directional identifier to distinguish the successively corresponding reasoning path of each order in the query process; and

any two of the multi-level directional identifiers have at least one of different colors, sizes, and shapes.

15 . The non-transitory computer-readable storage medium according to claim 12 , wherein the directional identifier is a directional arrow.

16 . A system, comprising:

at least one computing device comprising at least one hardware processor; and

a memory storing program instructions executable by the at least one hardware processor that, when executed, direct the at least one computing device to implement a human-computer interaction method, comprising:

receiving, by a terminal device, an input question;

extracting, by a terminal device, entities and relationships involved in the input question;

querying, by a server, an answer to the input question in a knowledge graph according to the entities and the relationships;

displaying, by a terminal device, a sub-graph of the knowledge graph, wherein the sub-graph of the knowledge graph comprises entities involved in the input question and the answer, and a directional identifier, wherein the directional identifier is configured to identify a reasoning path corresponding to the query process; and

in an instance in which the sub-graph of the knowledge graph is displayed by the terminal device:

detecting a degree of association between a current input question and a previous dialog entity, and determining whether to delete a previous sub-graph of the knowledge graph based on the degree of association;

in response to determining that the current input question and a previous dialog entity are not relevant, deleting the previous sub-graph of the knowledge graph, obtaining and displaying a new sub-graph of the knowledge graph corresponding to the current input question; and

in response to determining that the current input question and a previous dialog entity are relevant, continuing to add entities and relationships to the previous sub-graph of the knowledge graph,

wherein, in the instance in which the sub-graph of the knowledge graph is displayed by the terminal device, the method further comprises:

acquiring at least one of user attribute data, user behavior data, and operation data, wherein the user attribute data comprises at least one of age, gender, and purchasing power of a user, and the user behavior data comprises historical search data;

acquiring a recommended object according to at least one of the user attribute data, the user behavior data, and the operation data; and

displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph,

wherein displaying differentially the recommended object relative to the entity in the sub-graph of the knowledge graph comprises:

popping up a message prompt window in response to an operation of a user of clicking the entity, and displaying the recommended object in the message prompt window.

17 . The system according to claim 16 , wherein the directional identifier is configured to connect entities that are successively went through in the query process.

18 . The system according to claim 17 , wherein:

in response to the input question being a multi-order reasoning question, the directional identifier comprises a multi-level directional identifier to distinguish the successively corresponding reasoning path of each order in the query process; and

any two of the multi-level directional identifiers have at least one of different colors, sizes, and shapes.

19 . The system according to claim 16 , wherein the directional identifier is a directional arrow.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (34)