Human-computer interaction method and system

Abstract

The invention discloses a man-machine interaction method which comprises the following steps: executing to a current node, broadcasting a talking skill of the node by a robot, and making a response after a customer receives the talking skill; when the client interrupts the dialogue in the man-machine interaction process, the robot judges whether the currently executed node is allowed to be interrupted or not; when the node does not allow interruption, the robot continuously broadcasts the verbal skill until the end, and waits for the customer to respond or end; when the node allows interruption, the robot identifies whether an interruption condition is hit; when the target is hit, the robot pauses the current verbal skill broadcast and turns to a derivative branch; if not, the robot continues to carry out verbal skill broadcasting until the end; and when the transferred derivative branch node process is finished, returning to the original node before jumping, executing the original node return configuration, and waiting for the customer to respond or finish the process. In addition, the invention further discloses a man-machine interaction system. According to the method, on the basis of node interruption configuration, the interruption intention of a client is judged and recognized in combination with semantics and key operation, and the man-machine interaction experience is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of human-computer interaction, and in particular to a human-computer interaction method and system supporting conversation interruption. Background Art

[0002] In the interactive voice response (IVR) system of the prior art, the robot usually does not support the interruption of the human-computer dialogue. The robot needs to start accepting the user's response after the node broadcast is finished, which makes the human-computer interaction process relatively rigid and cannot well meet the emotional needs of customers.

[0003] Even if the robot supports conversation interruptions, it does not have the ability to judge the customer's intention to interrupt, and it is easy to misjudge the customer's responsive answers such as "um" and "yes" as interruptions. When the robot ends the interruption and derives the conversation and returns to the original node, it usually restarts the original node process from the starting point. The customer needs to repeat the same previous conversation process, resulting in low interaction fluency during the human-computer dialogue and poor user experience. Summary of the invention

[0004] Based on this, in order to solve the technical problems in the prior art, a human-computer interaction method is proposed, including:

[0005] Start the human-machine interaction process, and the robot interacts with the customer;

[0006] The human-computer interaction process is composed of one or more nodes and the logical configuration within the nodes; each node is provided with a corresponding speech script for the robot to broadcast to the customer; one or more nodes in the human-computer dialogue process are respectively set as nodes that do not allow interruption or nodes that allow interruption;

[0007] When the human-computer interaction process reaches the current node, the robot starts to broadcast the words of the node, and the customer responds after receiving the robot's words; when the customer interrupts the conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption;

[0008] When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method;

[0009] When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends;

[0010] When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method.

[0011] In one embodiment, when the customer utters consulting language or emotional language during the response process or the customer performs a key operation, the robot determines that the customer has an intention to interrupt;

[0012] For nodes that allow interruption, the robot can only interrupt and switch to the derivative branch when it determines that the customer has the intention to interrupt, that is, the customer's conversation content contains consulting language or emotional language keywords, or the customer presses a button; otherwise, when the robot determines that the user has only made a simple responsive response and has no intention to interrupt, the current speech broadcast will continue.

[0013] In one embodiment, the robot searches the interruption word library connected to it, and determines whether the conversation content issued by the customer hits the keywords in the interruption word library, or confirms whether the customer has performed a key operation;

[0014] When confirming that the customer's conversation content hits the keywords in the interruption vocabulary, the robot identifies the customer's intention to interrupt through the keywords;

[0015] When confirming that the customer has performed a key operation, the robot waits for the customer to finish keying and recognizes the customer's intention to interrupt through the key operation;

[0016] The robot jumps into the derivative branch according to the recognized intention, conducts human-computer interaction in the derived branch node, and broadcasts the words of the derived branch node.

[0017] In one embodiment, for a node set to a conversation interaction mode or a key mode, whether to allow interruption and interruption conditions are pre-configured; the attributes of the interruption conditions are configured in the node, including:

[0018] Configure the interactive function mode for interruption, including conversation interaction mode and key mode;

[0019] Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key.

[0020] Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words.

[0021] In one embodiment, one or more nodes in the human-computer dialogue process are set as nodes that do not allow interruption; for nodes that are set as not allowing interruption, any form of interruption will not stop the robot's speech broadcast at the current node;

[0022] Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption;

[0023] Among them, intent is the key meaning of the customer's conversation content in the interaction; the robot includes a recognition engine, which reads the customer's conversation and identifies the conversation content, associates the customer's conversation content and intent, and uses the identified intent as the basis for the robot's branch jump.

[0024] In addition, in order to solve the technical problems in the prior art, a human-computer interaction system is proposed, including a robot and an interruption vocabulary; the robot is connected to the interruption vocabulary;

[0025] The human-machine interaction system starts the human-machine interaction process, and the robot interacts with the customer;

[0026] The human-computer interaction process is composed of one or more nodes and the logical configuration within the nodes; each node is provided with a corresponding speech script for the robot to broadcast to the customer; one or more nodes in the human-computer dialogue process are respectively set as nodes that do not allow interruption or nodes that allow interruption;

[0027] When the human-computer interaction process reaches the current node, the robot starts to broadcast the words of the node, and the customer responds after receiving the robot's words; when the customer interrupts the conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption;

[0028] When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method;

[0029] When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends;

[0030] When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method.

[0031] In one embodiment, for a node set to a conversation interaction mode or a key mode, whether to allow interruption and interruption conditions are pre-configured; the attributes of the interruption conditions are configured in the node, including:

[0032] Configure the interactive function mode for interruption, including conversation interaction mode and key mode;

[0033] Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key.

[0034] Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words.

[0035] In one embodiment, one or more nodes in the human-computer dialogue process are set as nodes that do not allow interruption; for nodes that are set as not allowing interruption, any form of interruption will not stop the robot's speech broadcast at the current node;

[0036] Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption;

[0037] Among them, intent is the key meaning of the customer's conversation content in the interaction; the robot includes a recognition engine, which reads the customer's conversation and identifies the conversation content, associates the customer's conversation content and intent, and uses the identified intent as the basis for the robot's branch jump.

[0038] In one embodiment, after the robot completes the branch jump derived from the customer's interruption, it returns to the node before the interruption and sets the operation after the return in the original node return configuration, including updating the assignment, triggering SMS, and broadcasting the jump script.

[0039] In one embodiment, the interruption vocabulary is generated by the human-computer interaction system after the robot is created, and then the interruption vocabulary is maintained according to business scenarios and user needs. The user of the human-computer interaction system customizes the fields to add keywords, thereby updating the interruption vocabulary.

[0040] Implementing the embodiments of the present invention will have the following beneficial effects:

[0041] The present invention proposes a node-based interruption configuration based on the human-computer collaborative dialogue system, combines semantics and key operation judgment, identifies the customer's interruption intention to interrupt the current node broadcast and switch to a derivative branch, returns to the original node and then performs return processing, thereby achieving a smoother and more realistic human-computer interaction experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] in:

[0044] Figure 1 A schematic diagram of the flow of the human-computer interaction method of the present invention;

[0045] Figure 2 is a schematic diagram of a human-computer interaction system of the present invention; DETAILED DESCRIPTION

[0046] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] The present invention discloses a human-computer interaction method, comprising:

[0048] Start the human-machine interaction process, and the robot interacts with the customer;

[0049] The human-computer interaction process consists of one or more nodes and the logical configuration within the nodes; each node is equipped with a corresponding speech that the robot broadcasts to the customer;

[0050] Among them, for nodes set to conversation interaction mode or key mode, whether to allow interruption and the interruption conditions are pre-configured;

[0051] Specifically, the properties of the interruption condition are configured in the node, including:

[0052] Configure the interactive function mode for interruption, including conversation interaction mode and key mode;

[0053] Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key.

[0054] Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words;

[0055] In particular, in the dialogue interaction mode node, the recognition model, robot response speed, and customer input-free time are set; in the key mode node, the key type and node end method are set;

[0056] One or more nodes in the human-computer dialogue process are respectively set as nodes that allow interruption or nodes that allow interruption;

[0057] Among them, one or more nodes in the human-computer dialogue process are set as nodes that do not allow interruptions. For nodes that are set as not allowing interruptions, any form of interruption will not stop the robot's speech broadcast at the current node. Since interruptions are not allowed, important information on the node can be effectively and clearly conveyed to customers.

[0058] For example, in the human-computer dialogue process of the financial investment business background, when a node involves a risk warning of an investment product, the node is set as a node that does not allow interruption;

[0059] Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption;

[0060] Among them, the intention is the key meaning of the customer's conversation content during the interaction; the robot's recognition engine reads the customer's conversation and identifies the conversation content, associates the customer's conversation content with the intention, and uses the recognized intention as the basis for the robot's branch jump;

[0061] The human-computer interaction process is executed to the current node. The robot starts to broadcast the speech of this node. The customer responds after receiving the robot's speech.

[0062] When a customer interrupts a conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption;

[0063] Specifically, when the customer utters consulting language or emotional language during the response process, or the customer performs a key operation, the robot determines that the customer has the intention to interrupt;

[0064] Specifically, for nodes that allow interruption, the robot can only interrupt and switch to the derivative branch when it determines that the customer has the intention to interrupt, that is, the customer's conversation content contains consulting language or emotional language keywords, or the customer presses a button; otherwise, when the robot determines that the user has only made a simple responsive response (such as "um", "yes") and has no intention to interrupt, the current speech broadcast will continue;

[0065] When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method;

[0066] When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends;

[0067] Specifically, the robot searches the interruption vocabulary connected to it and determines whether the conversation content issued by the customer hits the keywords in the interruption vocabulary, or confirms whether the customer has performed a key operation;

[0068] In particular, the interruption vocabulary is generated by the human-computer interaction system after the robot is created. It is then maintained according to business scenarios and user needs, and the user customizes the fields for adding keywords to update the interruption vocabulary.

[0069] When confirming that the customer's conversation content hits the keywords in the interruption vocabulary, the robot identifies the customer's intention to interrupt through the keywords;

[0070] When confirming that the customer has performed a key operation, the robot waits for the customer to finish keying and recognizes the customer's intention to interrupt through the key operation;

[0071] The robot jumps to the derived branch according to the recognized intention, performs human-computer interaction in the derived branch node, and broadcasts the words of the derived branch node;

[0072] When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method;

[0073] In particular, after the robot completes the branch jump derived from the customer's interruption, it returns to the node before the interruption and sets the post-return operation in the original node return configuration, including updating the assignment, triggering SMS, and broadcasting the jump script; the original node return configuration operation improves the realism of human-computer interaction.

[0074] The present invention discloses a human-computer interaction system, comprising a robot and an interruption word library; the robot is connected to the interruption word library;

[0075] The human-machine interaction system starts the human-machine interaction process, and the robot interacts with the customer;

[0076] The human-computer interaction process consists of one or more nodes and the logical configuration within the nodes; each node is equipped with a corresponding speech that the robot broadcasts to the customer;

[0077] Among them, for nodes set to conversation interaction mode or key mode, whether to allow interruption and the interruption conditions are pre-configured;

[0078] Specifically, the properties of the interruption condition are configured in the node, including:

[0079] Configure the interactive function mode for interruption, including conversation interaction mode and key mode;

[0080] Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key.

[0081] Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words;

[0082] In particular, in the dialogue interaction mode node, the recognition model, robot response speed, and customer input-free time are set; in the key mode node, the key type and node end method are set;

[0083] Among them, one or more nodes in the human-computer dialogue process are respectively set as nodes that allow interruption or nodes that allow interruption;

[0084] Among them, one or more nodes in the human-computer dialogue process are set as nodes that do not allow interruptions. For nodes that are set as not allowing interruptions, any form of interruption will not stop the robot's speech broadcast at the current node. Since interruptions are not allowed, important information on the node can be effectively and clearly conveyed to customers.

[0085] For example, in the human-computer dialogue process of the financial investment business background, when a node involves a risk warning of an investment product, the node is set as a node that does not allow interruption;

[0086] Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption;

[0087] Among them, the intention is the key meaning of the customer's conversation content during the interaction; the robot includes a recognition engine, which reads the customer's conversation and then recognizes the conversation content, associates the customer's conversation content with the intention, and uses the recognized intention as the basis for the robot's branch jump;

[0088] The human-computer interaction process is executed to the current node. The robot starts to broadcast the speech of this node. The customer responds after receiving the robot's speech.

[0089] When a customer interrupts a conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption;

[0090] Specifically, when the customer utters consulting language or emotional language during the response process, or the customer performs a key operation, the robot determines that the customer has the intention to interrupt;

[0091] Specifically, for nodes that allow interruption, the robot can only interrupt and switch to the derivative branch when it determines that the customer has the intention to interrupt, that is, the customer's conversation content contains consulting language or emotional language keywords, or the customer presses a button; otherwise, when the robot determines that the user has only made a simple responsive response (such as "um", "yes") and has no intention to interrupt, the current speech broadcast will continue;

[0092] When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method;

[0093] When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends;

[0094] Specifically, the robot searches the interruption vocabulary connected to it and determines whether the conversation content issued by the customer hits the keywords in the interruption vocabulary, or confirms whether the customer has performed a key operation;

[0095] In particular, the interruption word library is generated by the human-computer interaction system after the robot is created, and then the interruption word library is maintained according to business scenarios and user needs. The user of the human-computer interaction system customizes the fields to add keywords, thereby updating the interruption word library;

[0096] When confirming that the customer's conversation content hits the keywords in the interruption vocabulary, the robot identifies the customer's intention to interrupt through the keywords;

[0097] When confirming that the customer has performed a key operation, the robot waits for the customer to finish keying and recognizes the customer's intention to interrupt through the key operation;

[0098] The robot jumps to the derived branch according to the recognized intention, performs human-computer interaction in the derived branch node, and broadcasts the words of the derived branch node;

[0099] When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method;

[0100] In particular, after the robot completes the branch jump derived from the customer's interruption, it returns to the node before the interruption and sets the post-return operation in the original node return configuration, including updating the assignment, triggering SMS, and broadcasting the jump script; the original node return configuration operation improves the realism of human-computer interaction.

[0101] Implementing the embodiments of the present invention will have the following beneficial effects:

[0102] The present invention can realize whether interruption is allowed according to the node setting in the human-computer interaction system, and combine intention recognition, semantic understanding and key discrimination to more accurately identify whether the customer response has the intention to interrupt before executing the interruption, which greatly improves the configuration flexibility and realism of human-computer interaction.

[0103] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A human-computer interaction method, characterized in that: include: Start the human-machine interaction process, and the robot interacts with the customer; The human-computer interaction process is composed of one or more nodes and the logical configuration within the nodes; each node is provided with a corresponding speech script for the robot to broadcast to the customer; one or more nodes in the human-computer dialogue process are respectively set as nodes that do not allow interruption or nodes that allow interruption; When the human-computer interaction process reaches the current node, the robot starts to broadcast the words of the node, and the customer responds after receiving the robot's words; when the customer interrupts the conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption; When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method; When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends; When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method.

2. The human-computer interaction method according to claim 1, characterized in that: in, When the customer uses consulting language or emotional language during the response process, or the customer performs a key operation, the robot determines that the customer has the intention to interrupt; For nodes that allow interruption, the robot can only interrupt and switch to the derivative branch when it determines that the customer has the intention to interrupt, that is, the customer's conversation content contains consulting language or emotional language keywords, or the customer presses a button; otherwise, when the robot determines that the user has only made a simple responsive response and has no intention to interrupt, the current speech broadcast will continue.

3. The human-computer interaction method according to claim 2, It is characterized in that in, The robot searches the interruption vocabulary connected to it and determines whether the conversation content issued by the customer hits the keywords in the interruption vocabulary, or confirms whether the customer has performed a key operation; When confirming that the customer's conversation content hits the keywords in the interruption vocabulary, the robot identifies the customer's intention to interrupt through the keywords; When confirming that the customer has performed a key operation, the robot waits for the customer to finish keying and recognizes the customer's intention to interrupt through the key operation; The robot jumps into the derivative branch according to the recognized intention, conducts human-computer interaction in the derived branch node, and broadcasts the words of the derived branch node.

4. The human-computer interaction method according to claim 1, It is characterized in that Among them, for a node set to a conversation interaction mode or a key mode, whether to allow interruption and the interruption conditions are pre-configured; the properties of the interruption conditions are configured in the node, including: Configure the interactive function mode for interruption, including conversation interaction mode and key mode; Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key. Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words.

5. The human-computer interaction method according to claim 1, characterized in that: in, Set one or more nodes in the human-machine dialogue process as nodes that do not allow interruption; for nodes that are set as not allowing interruption, any form of interruption will not stop the robot's speech broadcast at the current node; Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption; Among them, intent is the key meaning of the customer's conversation content in the interaction; the robot includes a recognition engine, which reads the customer's conversation and identifies the conversation content, associates the customer's conversation content and intent, and uses the identified intent as the basis for the robot's branch jump.

6. A human-computer interaction system, characterized in that: Including a robot and an interruption vocabulary; the robot is connected to the interruption vocabulary; The human-machine interaction system starts the human-machine interaction process, and the robot interacts with the customer; The human-computer interaction process is composed of one or more nodes and the logical configuration within the nodes; each node is provided with a corresponding speech script for the robot to broadcast to the customer; one or more nodes in the human-computer dialogue process are respectively set as nodes that do not allow interruption or nodes that allow interruption; When the human-computer interaction process reaches the current node, the robot starts to broadcast the words of the node, and the customer responds after receiving the robot's words; when the customer interrupts the conversation during the human-computer interaction process, the robot determines whether the currently executing node allows interruption; When the node is judged as a node that does not allow interruption, the robot continues to broadcast the speech of the node until the end of the broadcast, waiting for the customer's response or ending the current node according to the node setting method; When judging that a node is allowed to be interrupted, the robot identifies whether it meets the set interruption condition; when the interruption condition is met, the robot pauses the speech broadcast of the current node and switches to the derived branch for human-computer interaction; when the interruption condition is not met, the robot continues to broadcast the speech of the current node until the current human-computer interaction process ends; When the derived branch node process ends, the robot returns the current process to the original node before the jump and executes the original node return configuration, waiting for the customer's response or ending the current human-computer interaction process according to the original node setting method.

7. The human-computer interaction system according to claim 6, It is characterized in that Among them, for a node set to a conversation interaction mode or a key mode, whether to allow interruption and the interruption conditions are pre-configured; the properties of the interruption conditions are configured in the node, including: Configure the interactive function mode for interruption, including conversation interaction mode and key mode; Configure whether to enable conversation interruption and / or key interruption. For nodes set to conversation interaction mode, conversation interruption and key interruption can be selected only when they are also set to key mode. For key mode nodes, you can only select whether to allow customers to interrupt by key. Configure the minimum duration and minimum number of words, that is, the current node needs to be interrupted only after the robot has completed the minimum broadcast duration or the minimum number of words.

8. The human-computer interaction system according to claim 6, characterized in that: in , set one or more nodes in the human-computer dialogue process as nodes that do not allow interruption; for nodes that are set to not allow interruption, any form of interruption will not stop the robot's speech broadcast at the current node; Among them, one or more nodes in the human-computer dialogue process are set as nodes that allow interruption; for the nodes set to allow interruption, when the node is in dialogue interaction mode, the robot judges and recognizes the customer's intention to interrupt by searching for keywords in the interruption vocabulary connected to it; when the node is in key mode, the customer's key operation is recognized as interruption; when the node is set to dialogue interaction mode and key mode at the same time, the customer's key operation is also recognized as interruption; Among them, intent is the key meaning of the customer's conversation content in the interaction; the robot includes a recognition engine, which reads the customer's conversation and identifies the conversation content, associates the customer's conversation content and intent, and uses the identified intent as the basis for the robot's branch jump.

9. The human-computer interaction system according to claim 6, characterized in that: in, After the robot completes the branch jump derived from the customer's interruption, it returns to the node before the interruption and sets the post-return operation in the original node return configuration, including updating the assignment, triggering SMS, and broadcasting the jump script.

10. The human-computer interaction system according to claim 6, characterized in that: in, The interruption vocabulary is generated by the human-computer interaction system after the robot is created. It is then maintained according to business scenarios and user needs. Users of the human-computer interaction system customize the fields by adding keywords to update the interruption vocabulary.

Images