A dialogue method, device, equipment and medium for pre-filling dialogue node parameters

Abstract

The present invention proposes a dialogue method, device, equipment and medium for pre-filling dialogue node parameters, which relates to the field of artificial intelligence. The method comprises: an interactive device starts a dialogue, outputs an inquiry statement corresponding to a current dialogue node, and obtains user input information corresponding to the inquiry statement; the user input information is compared with the intention parameters of all dialogue nodes pre-stored by the interactive device, and it is determined according to the comparison result whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information. If so, the intention parameter of the corresponding dialogue node to be filled is loaded into the current dialogue node and associated with the corresponding statement in the input information; after completing the question and answer of the current dialogue node, the dialogue node is jumped, and it is determined whether the dialogue node after the jump is the dialogue node to be filled. If so, the inquiry of the dialogue node to be filled is skipped and the next dialogue node is entered until the dialogue process is completed.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and in particular to a dialogue method, device, equipment and medium for pre-filling dialogue node parameters. Background Art

[0002] Existing conversational robots are basically configured through process templates. During the conversation, the robot will conduct the conversation according to the script configured in the template and identify the customer's intention, so as to find out what to say next on the template. Therefore, the entire conversation process is carried out sequentially on the template.

[0003] However, the existing template conversation robot has a flaw. If the customer says some useful information in advance, but the information is not recognized on the current template node, the robot will not record it. When it reaches the next node, it will still ask the customer again to obtain the necessary information. The robot lacks memory of user input information, and the conversation process is too rigid and lacks flexibility, resulting in a poor user experience. Summary of the invention

[0004] In view of the above problems existing in the prior art, the present invention proposes a dialogue method, device, equipment and medium for pre-filling dialogue node parameters, which mainly solves the problem that the existing solution lacks flexibility in the intention dialogue process through speech templates, resulting in poor user experience.

[0005] In order to achieve the above-mentioned purpose and other purposes, the technical solution adopted by the present invention is as follows.

[0006] A dialog method for pre-filling dialog node parameters, comprising:

[0007] The interactive device starts a dialogue, outputs a query statement corresponding to the current dialogue node, and obtains user input information corresponding to the query statement;

[0008] Compare the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, and determine whether there is a dialog node to be filled other than the current dialog node that matches the user input information according to the comparison result; if so, load the corresponding intention parameter of the dialog node to be filled into the current dialog node and associate it with the corresponding sentence in the input information;

[0009] After completing the question and answer of the current dialogue node, the dialogue node jumps to determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

[0010] Optionally, the interactive device starts a dialogue, including:

[0011] Obtaining a user's biometric features, obtaining the user's identity information based on the biometric features, and initiating a conversation based on the identity information; and / or,

[0012] The business requirement information input by the user is obtained, a corresponding business session process is matched according to the business requirement information, and a session is started according to the business session process.

[0013] Optionally, before comparing the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, the following steps are included:

[0014] Acquire the intent parameters contained in all the dialog nodes of the interactive device, and construct an intent parameter table of the corresponding service according to the intent parameters, wherein the intent parameters include the intent of the dialog node and the identification information of the dialog node;

[0015] Inputting the user input information obtained by the current dialog node into a preset intention recognition model to obtain the user intention;

[0016] Acquire matching intent parameters from the intent parameter table according to the user intent.

[0017] Optionally, a method of obtaining the intent recognition model includes:

[0018] Obtain synonyms or near-synonyms of each dialogue intent to construct a training sample set;

[0019] Construct a neural network framework, input the training sample set into the neural network framework, adjust the neural network parameters, and obtain the intent recognition model.

[0020] Optionally, judging whether there is a to-be-filled dialog node other than the current dialog node that matches the user input information according to the comparison result, and if so, loading the intent parameter of the corresponding to-be-filled dialog node into the current dialog node and associating it with the corresponding sentence in the input information, including:

[0021] Determine whether the dialog node to be filled that matches the user input information in the current dialog node is a previous dialog node. If it is a previous dialog node, jump to the previous dialog node and obtain the intent parameter loaded in the previous dialog node.

[0022] Determine whether the intention parameter loaded by the previous dialog node corresponds to the dialog node between the previous dialog node and the current dialog node, record it as an intermediate node, skip the query statement of the intermediate node, and execute a node jump from the previous dialog node to the current dialog node.

[0023] Optionally, after obtaining the user input information corresponding to the query statement, the method further includes:

[0024] Acquire a user input statement associated with the intention parameter for storage, and acquire a query statement of the corresponding dialog node to be filled according to the intention parameter of the dialog node to be filled, and output the query statement and the associated corresponding user input statement to a preset interactive interface for information display;

[0025] After completing the dialogue process of the corresponding business, the user's business needs are obtained based on the user input information contained in each jump dialogue node and the user input statements associated with the intent parameters loaded by each dialogue node.

[0026] Optionally, obtain the user's business requirement output, including:

[0027] Matching preset business rules according to the jump logic of the dialog node and the loaded intent parameters;

[0028] The corresponding business response data is obtained from the business database according to the business rules.

[0029] A dialog device with pre-filled dialog node parameters, comprising:

[0030] A wake-up module is used for the interactive device to start a dialogue, output an inquiry statement corresponding to the current dialogue node, and obtain user input information corresponding to the inquiry statement;

[0031] a parameter pre-filling module, configured to compare the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, and determine whether there is a dialog node to be filled other than the current dialog node that matches the user input information according to the comparison result; if so, load the intention parameter of the corresponding dialog node to be filled into the current dialog node and associate it with the corresponding sentence in the input information;

[0032] The node jump module is used to jump the dialogue node after completing the question and answer of the current dialogue node, and determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

[0033] A computer device comprises: a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the dialog method for pre-filling dialog node parameters when executing the computer program.

[0034] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the dialog method for pre-filling dialog node parameters are implemented.

[0035] As described above, the dialogue method, device, equipment and medium for pre-filling dialogue node parameters of the present invention have the following beneficial effects.

[0036] By comparing the user's inquiry statement with the intent parameters of all dialogue nodes, the intent parameters of other dialogue nodes corresponding to the inquiry statement are preloaded into the current dialogue node, and the preloaded intent parameters are associated with the corresponding statements in the inquiry statement. The user input information is recorded. When the dialogue node jumps, the inquiry process of the preloaded dialogue node to be filled can be skipped, which simplifies the dialogue process and enhances the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 The figure is a flowchart of a dialog method for pre-filling dialog node parameters in one embodiment of the present invention.

[0038] Figure 2 The figure is a flow chart of initiating a dialogue by an interactive device in one embodiment of the present invention.

[0039] Figure 3 The figure is a flow chart of intent parameter matching of dialog nodes in one embodiment of the present invention.

[0040] Figure 4 A module diagram of a dialog device for pre-filling dialog node parameters in an embodiment of the present invention.

[0041] Figure 5 FIG. 1 is a schematic diagram of the structure of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0042] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.

[0043] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0044] The inventors have found that for business robots such as robots that review and approve loans, when the conversation reaches the second node, the robot asks the customer: "Do you have a house under your name?" The customer answers: "I don't have a house, but I have a car." The customer's words actually express two intentions, namely "I don't have a house" and "I have a car." However, although the robot recognizes these two intentions, for this node, the robot will only select "I don't have a house" for processing and ignore the intention of "I have a car." Then when the conversation reaches the third node, according to the template configuration, the robot will ask the customer again: "Then do you have a car under your name?" This causes the customer to be confused due to repeated questions during the conversation. It is inferred that it is not a bank employee, but a robot that is talking to him. This lack of memory interaction is not humanistic and not very friendly.

[0045] See also Figure 1 In view of the problems existing in the robot dialogue process, the present invention provides a dialogue method, device, equipment and medium for pre-filling dialogue node parameters, including the following steps:

[0046] S1: The interactive device starts a dialogue, outputs a query statement corresponding to the current dialogue node, and obtains user input information corresponding to the query statement;

[0047] S2: Compare the user input information with the intention parameters of all dialog nodes pre-stored in the interactive device to determine whether there is a dialog node to be filled other than the current dialog node that matches the user input information; if so, load the corresponding intention parameter of the dialog node to be filled into the current dialog node and associate it with the corresponding sentence in the input information;

[0048] S3: After completing the question and answer of the current dialogue node, the dialogue node is jumped to determine whether the dialogue node after the jump is the dialogue node to be filled. If so, the inquiry of the corresponding dialogue node is skipped and the next dialogue node is entered until the dialogue process is completed.

[0049] The dialog method for pre-filling dialog node parameters provided by the present invention is described below in conjunction with specific embodiments.

[0050] In step S1, the interactive device starts a dialogue, outputs a query statement corresponding to a current dialogue node, and obtains user input information corresponding to the query statement.

[0051] See also Figure 2 In one embodiment, the interactive device initiates a conversation, including the following steps:

[0052] Step S101, obtaining a user's biometric features, obtaining the user's identity information based on the biometric features, and initiating a conversation based on the identity information; and / or,

[0053] Step S102: acquiring the service requirement information input by the user, matching the corresponding service session flow according to the service requirement information, and starting the session according to the service session flow.

[0054] Specifically, the interactive device may include a question-and-answer robot, a portable interactive terminal, or a terminal device with a display interface, etc. Taking the question-and-answer robot as an example, a collection device for collecting user biometric features may be provided on the question-and-answer robot. Among them, biometric features include: fingerprint features, facial features, iris features, voice features, etc. The question-and-answer robot may select one or more of these biometric features for user identity recognition. Exemplarily, the collection device may use a face recognition module. After the question-and-answer robot collects a face image through a camera, it inputs the face image into the face recognition module, obtains the user identity information corresponding to the face by recognizing the face image, and then starts a conversation based on the identity information. If the face recognition user is "Wang XX", the question-and-answer robot may output "Hello, Mr. Wang! How can I help you?", and outputs a query statement for a specific object based on the identity information.

[0055] It is also possible to directly wake up the conversation of the interactive device according to the business demand information input by the user without distinguishing the user identity. Specifically, the business conversation process corresponding to different businesses can be pre-set for different businesses, the conversation node of the specific business can be determined, and the corresponding business conversation process can be represented based on the business category. The user can input the business demand information through voice or text. After the interactive device recognizes the business demand information in the voice or text, it matches the corresponding business category according to the business demand information, and obtains the business conversation process corresponding to the business category. The similarity between the features corresponding to the business demand information and the business category can be calculated by using the paradigm distance or the Euclidean distance, and whether the two match can be determined according to the similarity threshold. In another embodiment, the basic conversation of the interactive device can also be woken up by the biometric feature, such as asking the user's needs, "Mr. xx, how can I help you?". The user's business demand information is obtained based on the basic conversation, and then the business demand information is matched with the corresponding business conversation process. The specific application and interaction method are not limited here.

[0056] It is also possible to determine whether the user has the query authority to request the service by collecting the user's biometric features, and then start the service session process of the corresponding service with the user's authority. The selection and setting of specific biometric features can be adjusted according to actual application needs and are not limited here.

[0057] In step S2, the user input information is compared with the intention parameters of all dialogue nodes pre-stored in the interactive device, and it is determined based on the comparison result whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information. If so, the intention parameters of the corresponding dialogue node to be filled are loaded into the current dialogue node and associated with the corresponding statement in the input information.

[0058] The business conversation process may include multiple conversation nodes, and connections or jump relationships between conversation nodes may be established according to conversation intent. Specifically, taking robot conversation as an example, conversation intent may include the following three categories:

[0059] (1) Global general intent: that is, the intention of objection, for example: wait a minute.

[0060] (2) Common intent: A common intent only corresponds to a question, that is, a common intent will only appear once in a robot. For example: I have a car.

[0061] (3) General process intention: Controls the direction of the process and will correspond to multiple questions. For example: yes, no.

[0062] Each dialog node can correspond to multiple conversation intentions. For example, a dialog node may include a common intention, multiple process common intentions, and a global common intention. During the conversation, when the user modifies the input information of the previous dialog node, the global common intention parameter of the current dialog node can be triggered through the global common intention such as "wait a moment" or "wait a moment", and the modified information can be input to jump to the dialog node where the error information was input previously. The process common intention is used to determine which connected lower-level dialog node the current dialog node needs to jump to. The common intention corresponds to the reply statement of the inquiry statement corresponding to the dialog node.

[0063] See also Figure 3 In one embodiment, before comparing the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, the following steps are included:

[0064] Step S103, acquiring the intent parameters contained in all the dialog nodes of the interactive device, and constructing an intent parameter table of the corresponding service according to the intent parameters, wherein the intent parameters include the intent of the dialog node and the identification information of the dialog node;

[0065] Step S104, inputting the user input information obtained by the current dialogue node into a preset intention recognition model to obtain the user intention;

[0066] Step S105: Acquire matching intent parameters from the intent parameter table according to the user intent.

[0067] Specifically, after the interactive device is awakened and determines the business conversation process, the dialog node corresponding to the business conversation process can be obtained, and then the intent parameters contained in each dialog node can be obtained to generate an intent parameter table containing all intent parameters in the business conversation process. Among them, each intent parameter contains the conversation intent of the dialog node, such as global general intent, process general intent, common intent, etc. Different dialog nodes can be distinguished by a unique identification code, and the unique identification code of the corresponding dialog node is stored in each intent parameter. After the intent parameter is determined, the dialog node where the intent parameter is located can be determined by the unique identification code.

[0068] In one embodiment, a hash value of each intent parameter may be calculated and stored in an intent parameter table to protect the intent data from being tampered with and improve data security. When comparing intent parameters, the data may be directly compared by hash value to determine whether they are consistent, which may also improve data comparison efficiency.

[0069] In one embodiment, the method of obtaining the intent recognition model includes:

[0070] Obtain synonyms or near-synonyms of each dialogue intent to construct a training sample set;

[0071] Construct a neural network framework, input the training sample set into the neural network framework, adjust the neural network parameters, and obtain the intent recognition model.

[0072] Specifically, the neural network framework can adopt the network model architecture of deep learning neural network and recurrent neural network to obtain the initial neural network framework. For the intent parameters contained in different business conversation processes, the sample data corresponding to each intent parameter is sorted, and the sample data includes synonyms, near synonyms, and cognates of the conversation intent. Exemplarily, the common intent is "I have a car", and the corresponding sample data is: "I have a sedan", "I have an Audi", etc. By constructing a sample training neural network framework, the neural network parameters are adjusted to obtain the corresponding relationship between the sample data and the intent parameters. The user input information can be directly mapped to the corresponding intent parameters through the intent recognition model obtained by training. The specific model training process is not repeated here. After obtaining the intent parameters corresponding to the user input information of the current dialogue node through the intent recognition model, the hash value of the acquired intent parameter can be calculated, and the corresponding intent parameter and the unique identification code of the corresponding dialogue node can be obtained from the intent parameter table according to the hash value. The user intent of the current dialogue node is obtained through the intent recognition model, and the intent parameters in the intent parameter table are matched according to the user intent. It can be determined whether the user input information of the current dialogue node corresponds to the intent parameters of other dialogue nodes, so that the corresponding intent parameters of other nodes can be loaded into the current dialogue node.

[0073] In one embodiment, judging whether there is a dialog node to be filled other than the current dialog node matching the user input information according to the comparison result, and if so, loading the intention parameter of the corresponding dialog node to be filled into the current dialog node and associating it with the corresponding sentence in the input information, including:

[0074] Determine whether the dialog node to be filled that matches the user input information in the current dialog node is a previous dialog node. If it is a previous dialog node, jump to the previous dialog node and obtain the intent parameter loaded in the previous dialog node.

[0075] Determine whether the intention parameter loaded by the previous dialog node corresponds to the dialog node between the previous dialog node and the current dialog node, record it as an intermediate node, skip the query statement of the intermediate node, and execute a node jump from the previous dialog node to the current dialog node.

[0076] Specifically, when the user input information of the current dialogue node matches the intention parameter of other dialogue nodes according to the aforementioned steps, the dialogue node to be filled and the corresponding intention parameter are determined according to the user input information in the current dialogue node, and the intention parameter of the current dialogue node is pre-filled. For example, if the current dialogue node interaction device outputs the query sentence "Do you have a house?", the user input information is "I have a house and a car", and "I have a car" corresponds to the intention parameter of another dialogue node, then the unique identification code of the dialogue node corresponding to "I have a car" and the dialogue node position can be determined by matching with the intention parameter table. Load the matching intention parameter to the current dialogue node, and store "I have a car" in the parameter value of the matching intention parameter to associate the intention parameter. A pre-fill record table can be set to record the pre-filled intention parameters of each dialogue node. After completing the entire dialogue process, the pre-fill record table can be directly queried to determine which dialogue nodes' intention parameters have been pre-filled.

[0077] In step S3, after completing the question and answer of the current dialogue node, the dialogue node jumps to determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

[0078] Specifically, after completing the question and answer of the current dialogue node, you can jump to the next dialogue node according to the jump relationship of the dialogue node, and you can determine whether the intention parameter of the next dialogue node has been pre-filled in the previous dialogue node. Specifically, you can determine the corresponding pre-filled intention parameter by querying the pre-filled record table, and then determine whether the intention parameter in the jumped dialogue node has been pre-filled in the previous dialogue node. If it has been pre-filled, skip the query statement of the corresponding dialogue node and determine the next jump dialogue node based on the user input statement associated with the pre-filled intention parameter, directly jump to the next dialogue node, and then determine whether the corresponding intention parameter has been pre-filled. If it has not been pre-filled, directly output the query statement of the dialogue node to obtain the user's input information, and then pre-fill the intention parameter based on the user input information.

[0079] When the user inputs a global general intent, the intent parameters of the previous dialog node can be pre-filled. For example, if the user input information in the current dialog node is "Wait a minute, I don't have a house in local A, but I have a house in B", and the intent parameter corresponding to "I have a house" corresponds to the dialog node before the current dialog node, then the current dialog node can jump to the corresponding previous dialog node. According to the user input sentence "I have a house in B" associated with the pre-filled intent parameter, the jump relationship of the dialog node is determined, and the intent parameters pre-filled in the previous dialog node according to the previous user input information are obtained. According to the pre-filled intent parameters of the previous dialog node, it is determined that the intent parameters of the dialog nodes between the previous dialog node and the current dialog node have been pre-filled, and these dialog nodes that have been pre-filled by the previous dialog node are recorded as intermediate nodes, and these intermediate nodes are skipped to simplify the jump back path from the previous dialog node to the current dialog node.

[0080] In one embodiment, after obtaining the user input information corresponding to the query statement, the method further includes:

[0081] Acquire a user input statement associated with the intention parameter for storage, and acquire a query statement of the corresponding dialog node to be filled according to the intention parameter of the dialog node to be filled, and output the query statement and the associated corresponding user input statement to a preset interactive interface for information display;

[0082] After completing the dialogue process of the corresponding business, the user's business needs are obtained based on the user input information contained in each jump dialogue node and the user input statements associated with the intent parameters loaded by each dialogue node.

[0083] Specifically, the query statements of the skipped dialogue nodes and the user input statements associated with the intention parameters of the skipped dialogue nodes can be displayed through the interactive interface, so that users can intuitively understand the information required by each dialogue node and grasp the demand information of the corresponding business. The next time the user uses it, he can adjust the input content of the dialogue node based on the business demand information he understands, improve the interaction efficiency, and further enhance the user experience.

[0084] In one embodiment, obtaining the user's business demand output includes:

[0085] Matching preset business rules according to the jump logic of the dialog node and the loaded intent parameters;

[0086] The corresponding business response data is obtained from the business database according to the business rules.

[0087] The correspondence between the transferred dialogue nodes and the business response data. For example, after dialogue nodes 1, 3, and 4, the corresponding business response data is A, and after dialogue nodes 1, 5, and 6, the corresponding business response data is B. The specific business rules can be set according to the actual application requirements and are not limited here.

[0088] Taking a robot as an example, suppose the robot has five nodes:

[0089] The first node asks if it is me;

[0090] The second node asks the customer if he has a room;

[0091] The third node asks the customer’s age;

[0092] The fourth node asks the customer if he has a car;

[0093] The fifth node is to determine whether the loan conditions are met based on the above inquiries and inform the customer of the loan amount that can be applied for.

[0094] Configure the routing intentions "I don't have a house" and "I have a house", jump to the second node, and configure the confirmation words.

[0095] Case process speech demonstration

[0096] Robot: Hello, I would like to ask you to confirm some matters regarding XX loan. Are you Mr. Wang?

[0097] Customer: Yes.

[0098] Robot: Hello, do you have a room?

[0099] Customer: I don’t have a house, but I have a car

[0100] (NLP returns two intents, "I don't have a house" and "I have a car". DM traverses preFillElementMap for the "I have a car" intent and finds that the intent is in the intentAndParam collection in the fourth node, so the following operation is performed: preFillElementMap.get("04").setValue("I have a car"))

[0101] Robot: How old are you?

[0102] Client: 30 years old.

[0103] (After age is assigned a value of 30, the program finds that preFillElementMap.get("04").getValue() has a value, so it processes the value "I have a car" as the intention of the customer's reply at the fourth node and skips the speech broadcast at the fourth node)

[0104] Robot: Hello, according to the review conditions, your loan limit is 20,000 yuan.

[0105] Customer: Sorry, I said the wrong thing before. I have a house in Shanghai.

[0106] (NLP returns the intent of "I have a house", DM triggers the routing intent jump, and the confirmation words are broadcast)

[0107] Robot: OK, are you sure we have a room?

[0108] Customer: Yes.

[0109] (After the customer confirms, according to the original logic, it will enter the third node to continue asking, but after the intention is pre-filled, preFillElementMap.get("03").getValue() and preFillElementMap.get t("04").getValue() both have values, so they are skipped one by one and enter the fifth node for broadcasting)

[0110] Robot: OK, based on the review criteria, your credit limit is 50,000 yuan.

[0111] In one embodiment, if Figure 4As shown, a dialogue device for pre-filling dialogue node parameters is provided, and the device includes: a wake-up module 10, which is used for an interactive device to start a dialogue, output an inquiry statement corresponding to a current dialogue node, and obtain user input information corresponding to the inquiry statement; a parameter pre-filling module 11, which is used to compare the user input information with the intention parameters of all dialogue nodes pre-stored by the interactive device, and determine whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information according to the comparison result. If so, the intention parameter of the corresponding dialogue node to be filled is loaded into the current dialogue node and associated with the corresponding statement in the input information; a node jump module 12, which is used to jump the dialogue node after completing the question and answer of the current dialogue node, and determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

[0112] In one embodiment, the wake-up module 10 includes: an identity authentication unit, used to obtain a user's biometric features, obtain the user's identity information based on the biometric features, and start a conversation based on the identity information; a business matching unit, used to obtain business demand information input by the user, match the corresponding business session process based on the business demand information, and start a session based on the business session process.

[0113] In one embodiment, the device also includes: a parameter table acquisition module, which is used to obtain the intention parameters contained in all dialogue nodes of the interactive device, and construct an intention parameter table of the corresponding business based on the intention parameters, wherein the intention parameters include the intention of the dialogue node and the identification information of the dialogue node; an intention recognition module, which is used to input the user input information obtained by the current dialogue node into a preset intention recognition model to obtain the user intention; and a parameter acquisition module, which obtains matching intention parameters from the intention parameter table based on the user intention.

[0114] In one embodiment, the parameter pre-filling module 11 includes: a parameter loading unit, which is used to determine whether the dialogue node to be filled that matches the user input information in the current dialogue node is a previous dialogue node. If it is a previous dialogue node, jump to the previous dialogue node and obtain the intention parameters loaded by the previous dialogue node; a node recording unit, which is used to determine whether the intention parameters loaded by the previous dialogue node correspond to the dialogue node between the previous dialogue node and the current dialogue node, record it as an intermediate node, skip the query statement of the intermediate node, and execute a node jump from the previous dialogue node to the current dialogue node.

[0115] In one embodiment, the device also includes: a dialogue information display module, which is used to obtain user input statements associated with the intent parameters for storage, and obtain the corresponding inquiry statements of the dialogue nodes to be filled according to the intent parameters of the dialogue nodes to be filled, and output the inquiry statements and the associated corresponding user input statements to a preset interactive interface for information display; a node data response module, which is used to obtain the user's business needs after completing the dialogue process of the corresponding business, according to the user input information contained in each jumped dialogue node, and the user input statements associated with the intent parameters loaded by each dialogue node.

[0116] In one embodiment, the node data response module includes: a business rule matching unit, which is used to match preset business rules according to the jump logic of the dialogue node and the loaded intent parameters; and a response output unit, which is used to obtain corresponding business response data from a business database according to the business rules.

[0117] In one embodiment, the tool matching unit includes: a card acquisition component, which is used to match cards according to the business needs, and display the acquired cards through the visual interactive interface, wherein the cards include the name and description information of the corresponding business management tool; and a selection output component, which is used to obtain the corresponding business management tool according to the card selected by the user.

[0118] The above-mentioned dialogue device for pre-filling dialogue node parameters can be implemented in the form of a computer program. The computer program can be implemented in the form of Figure 5 The computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor.

[0119] Each module in the dialogue device with pre-filled dialogue node parameters can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of the terminal's memory in the form of hardware, or can be stored in the terminal's memory in the form of software, so that the processor can call and execute the operations corresponding to each module. The processor can be a central processing unit (CPU), a microprocessor, a single-chip microcomputer, etc.

[0120] like Figure 5The figure shows a schematic diagram of the internal structure of a computer device in an embodiment. A computer device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program: an interactive device starts a dialogue, outputs an inquiry statement corresponding to a current dialogue node, and obtains user input information corresponding to the inquiry statement; the user input information is compared with the intention parameters of all dialogue nodes pre-stored by the interactive device, and it is determined whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information according to the comparison result, and if so, the intention parameter of the corresponding dialogue node to be filled is loaded into the current dialogue node and associated with the corresponding statement in the input information; after completing the question and answer of the current dialogue node, the dialogue node is jumped, and it is determined whether the dialogue node after the jump is the dialogue node to be filled, and if so, the inquiry of the dialogue node to be filled is skipped and the next dialogue node is entered until the dialogue process is completed.

[0121] In one embodiment, when the above-mentioned processor is executed, the interactive device implemented to initiate a conversation includes: obtaining a user's biometric features, obtaining the user's identity information based on the biometric features, and initiating a conversation based on the identity information; and / or obtaining business requirement information input by the user, matching a corresponding business session process based on the business requirement information, and initiating a conversation based on the business session process.

[0122] In one embodiment, when the above-mentioned processor is executed, before comparing the user input information with the intention parameters of all dialogue nodes pre-stored by the interactive device, the implementation includes: obtaining the intention parameters contained in all dialogue nodes of the interactive device, and constructing an intention parameter table of the corresponding business based on the intention parameters, wherein the intention parameters include the intention of the dialogue node and the identification information of the dialogue node; inputting the user input information obtained by the current dialogue node into a preset intention recognition model to obtain the user intention; and obtaining matching intention parameters from the intention parameter table based on the user intention.

[0123] In one embodiment, when the above-mentioned processor is executed, the method of obtaining the intent recognition model includes: obtaining synonyms or near-synonyms of each dialogue intention to construct a training sample set; constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intent recognition model.

[0124] In one embodiment, when the processor is executed, the implementation determines whether there is a dialog node to be filled other than the current dialog node that matches the user input information based on the comparison result. If so, the intent parameter of the corresponding dialog node to be filled is loaded into the current dialog node and associated with the corresponding statement in the input information, including: determining whether the dialog node to be filled that matches the user input information in the current dialog node is the previous dialog node. If it is the previous dialog node, jump to the previous dialog node and obtain the intent parameter loaded by the previous dialog node; determine whether the intent parameter loaded by the previous dialog node corresponds to the dialog node between the previous dialog node and the current dialog node, recorded as an intermediate node, skip the query statement of the intermediate node, and execute a node jump from the previous dialog node to the current dialog node.

[0125] In one embodiment, when the above-mentioned processor is executed, after obtaining the user input information corresponding to the query statement, it also includes: obtaining the user input statement associated with the intention parameter for storage, and obtaining the corresponding query statement of the dialogue node to be filled according to the intention parameter of the dialogue node to be filled, and outputting the query statement and the associated corresponding user input statement to a preset interactive interface for information display; after completing the dialogue process of the corresponding business, according to the user input information contained in each jumped dialogue node, and the user input statement associated with the intent parameter loaded by each dialogue node, the user's business needs are obtained.

[0126] In one embodiment, when the above-mentioned processor is executed, the obtained user's business demand output is implemented, including: matching the preset business rules according to the jump logic of the dialogue node and the loaded intent parameters; and obtaining the corresponding business response data from the business database according to the business rules.

[0127] In one embodiment, the above-mentioned computer device can be used as a server, including but not limited to an independent physical server, or a server cluster composed of multiple physical servers. The computer device can also be used as a terminal, including but not limited to a mobile phone, a tablet computer, a personal digital assistant or a smart device. Figure 5 As shown, the computer device includes a processor, a non-volatile storage medium, an internal memory, a display screen, and a network interface connected via a system bus.

[0128] Among them, the processor of the computer device is used to provide computing and control capabilities to support the operation of the entire computer device. The non-volatile storage medium of the computer device stores an operating system and a computer program. The computer program can be executed by the processor to implement the dialogue method for pre-filling dialogue node parameters provided in the above embodiments. The internal memory in the computer device provides a cache operating environment for the operating system and computer program in the non-volatile storage medium. The display interface can display data through a display screen. The display screen can be a touch screen, such as a capacitive screen or an electronic screen, and can generate corresponding instructions by receiving a click operation acting on a control displayed on the touch screen.

[0129] Those skilled in the art will understand that Figure 5 The structure of the computer device shown in the figure is only a block diagram of a partial structure related to the present application scheme, and does not constitute a limitation on the computer device to which the present application scheme is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0130] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented: an interactive device initiates a conversation, outputs an inquiry statement corresponding to a current conversation node, and obtains user input information corresponding to the inquiry statement; the user input information is compared with the intention parameters of all conversation nodes pre-stored by the interactive device, and it is determined based on the comparison result whether there is a conversation node to be filled other than the current conversation node that matches the user input information, and if so, the intention parameter of the corresponding conversation node to be filled is loaded into the current conversation node and associated with the corresponding statement in the input information; after completing the question and answer of the current conversation node, the conversation node is jumped, and it is determined whether the conversation node after the jump is the conversation node to be filled, and if so, the inquiry of the conversation node to be filled is skipped and the next conversation node is entered until the conversation process is completed.

[0131] In one embodiment, when the computer program is executed by a processor, the interactive device implemented to initiate a conversation includes: obtaining a user's biometric features, obtaining the user's identity information based on the biometric features, and initiating a conversation based on the identity information; and / or obtaining business requirement information input by the user, matching a corresponding business session process based on the business requirement information, and initiating a conversation based on the business session process.

[0132] In one embodiment, when the computer program is executed by a processor, before comparing the user input information with the intention parameters of all dialogue nodes pre-stored in the interactive device, the implementation includes: obtaining the intention parameters contained in all dialogue nodes of the interactive device, and constructing an intention parameter table of the corresponding business based on the intention parameters, wherein the intention parameters include the intention of the dialogue node and the identification information of the dialogue node; inputting the user input information obtained by the current dialogue node into a preset intention recognition model to obtain the user intention; and obtaining matching intention parameters from the intention parameter table based on the user intention.

[0133] In one embodiment, when the computer program is executed by a processor, the method of obtaining the intent recognition model includes: obtaining synonyms or near-synonyms of each dialogue intent to construct a training sample set; constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intent recognition model.

[0134] In one embodiment, when the computer program is executed by a processor, the implemented method is to determine whether there is a to-be-filled dialog node other than the current dialog node that matches the user input information based on the comparison result; if so, load the intent parameter of the corresponding to-be-filled dialog node into the current dialog node and associate it with the corresponding statement in the input information, including: determining whether the to-be-filled dialog node that matches the user input information in the current dialog node is a previous dialog node; if it is a previous dialog node, jump to the previous dialog node and obtain the intent parameter loaded by the previous dialog node; determine whether the intent parameter loaded by the previous dialog node corresponds to a dialog node between the previous dialog node and the current dialog node, recorded as an intermediate node, skip the query statement of the intermediate node, and execute a node jump from the previous dialog node to the current dialog node.

[0135] In one embodiment, when the computer program is executed by a processor, after obtaining the user input information corresponding to the query statement, it also includes: obtaining the user input statement associated with the intention parameter for storage, and obtaining the corresponding query statement of the dialogue node to be filled according to the intention parameter of the dialogue node to be filled, and outputting the query statement and the associated corresponding user input statement to a preset interactive interface for information display; after completing the dialogue process of the corresponding business, obtaining the user's business needs according to the user input information contained in each jumped dialogue node and the user input statement associated with the intention parameter loaded by each dialogue node.

[0136] In one embodiment, when the instruction is executed by the processor, the obtained user's business demand output includes: matching the preset business rules according to the jump logic of the dialogue node and the loaded intent parameters; and obtaining the corresponding business response data from the business database according to the business rules.

[0137] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM), etc.

[0138] In summary, the present invention provides a dialogue method, device, equipment and medium for pre-filling dialogue node parameters, and designs a mechanism for pre-filling the intention and parameters of robot dialogue nodes; by pre-filling the intention and parameters of the rear dialogue nodes of the current round, the problem of the robot's lack of memory of user input is solved, making the robot more intelligent in context understanding; if used together with routing intention jump and routing parameter jump, it can also realize upward modification and return to the original path, skipping the intermediate nodes that have not been modified, forming a closed loop of semantic understanding of the entire dialogue robot, which can understand in advance downward, and can also return to the current directly after upward modification, simplifying the dialogue process and enhancing user experience. Therefore, the present invention effectively overcomes the various shortcomings in the prior art and has a high industrial utilization value.

[0139] The above embodiments are merely illustrative of the principles and effects of the present invention, and are not intended to limit the present invention. Anyone familiar with the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, all equivalent modifications or alterations made by a person of ordinary skill in the art without departing from the spirit and technical concept disclosed by the present invention shall still be covered by the claims of the present invention.

Claims

1. A dialog method for pre-filling dialog node parameters, characterized in that: include: The interactive device starts a dialogue, outputs a query statement corresponding to the current dialogue node, and obtains user input information corresponding to the query statement; The user input information is compared with the intention parameters of all dialogue nodes pre-stored in the interactive device, and it is determined based on the comparison result whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information. If so, the intention parameter of the corresponding dialogue node to be filled is loaded into the current dialogue node and associated with the corresponding statement in the input information, including: determining whether the dialogue node to be filled that matches the user input information in the current dialogue node is a previous dialogue node, and if it is a previous dialogue node, jumping to the previous dialogue node and obtaining the intention parameter loaded in the previous dialogue node; determining whether the intention parameter loaded in the previous dialogue node corresponds to the previous dialogue node and the The dialog nodes between the current dialog nodes are recorded as intermediate nodes, the query statements of the intermediate nodes are skipped, and the node jump from the previous dialog node to the current dialog node is executed; before comparing the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, it includes: obtaining the intention parameters contained in all dialog nodes of the interactive device, and constructing an intention parameter table of the corresponding business according to the intention parameters, wherein the intention parameters include the intention of the dialog node and the identification information of the dialog node; inputting the user input information obtained by the current dialog node into a preset intention recognition model to obtain the user intention; obtaining the matching intention parameters from the intention parameter table according to the user intention; After completing the question and answer of the current dialogue node, the dialogue node jumps to determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

2. The dialog method for pre-filling dialog node parameters according to claim 1, characterized in that: The interactive device starts a dialogue, including: Obtaining a user's biometric features, obtaining the user's identity information based on the biometric features, and initiating a conversation based on the identity information; and / or, The business requirement information input by the user is obtained, a corresponding business session process is matched according to the business requirement information, and a session is started according to the business session process.

3. The dialog method for pre-filling dialog node parameters according to claim 1, characterized in that: The method of obtaining the intent recognition model includes: Obtain synonyms or near-synonyms of each dialogue intent to construct a training sample set; Construct a neural network framework, input the training sample set into the neural network framework, adjust the neural network parameters, and obtain the intent recognition model.

4. The dialog method for pre-filling dialog node parameters according to claim 1, characterized in that: After obtaining the user input information corresponding to the query statement, the method further includes: Acquire a user input statement associated with the intention parameter for storage, and acquire a query statement of the corresponding dialog node to be filled according to the intention parameter of the dialog node to be filled, and output the query statement and the associated corresponding user input statement to a preset interactive interface for information display; After completing the dialogue process of the corresponding business, the user's business needs are obtained based on the user input information contained in each jump dialogue node and the user input statements associated with the intent parameters loaded by each dialogue node.

5. The dialog method for pre-filling dialog node parameters according to claim 4, characterized in that: Obtain the user's business demand output, including: Matching preset business rules according to the jump logic of the dialog node and the loaded intent parameters; The corresponding business response data is obtained from the business database according to the business rules.

6. A dialogue device with pre-filled dialogue node parameters, characterized in that: include: A wake-up module is used for the interactive device to start a dialogue, output an inquiry statement corresponding to the current dialogue node, and obtain user input information corresponding to the inquiry statement; A parameter pre-filling module is used to compare the user input information with the intention parameters of all dialogue nodes pre-stored in the interactive device, and determine whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information based on the comparison result. If so, the intention parameter of the corresponding dialogue node to be filled is loaded into the current dialogue node and associated with the corresponding statement in the input information, including: determining whether the dialogue node to be filled that matches the user input information in the current dialogue node is a previous dialogue node, and if it is a previous dialogue node, jumping to the previous dialogue node and obtaining the intention parameter loaded in the previous dialogue node; determining whether the intention parameter loaded in the previous dialogue node corresponds to the previous dialogue node. The dialog node between the conversation node and the current dialog node is recorded as an intermediate node, the query statement of the intermediate node is skipped, and the node jump from the previous dialog node to the current dialog node is executed; before comparing the user input information with the intention parameters of all dialog nodes pre-stored by the interactive device, it includes: obtaining the intention parameters contained in all dialog nodes of the interactive device, and constructing an intention parameter table of the corresponding business according to the intention parameters, wherein the intention parameters include the intention of the dialog node and the identification information of the dialog node; inputting the user input information obtained by the current dialog node into a preset intention recognition model to obtain the user intention; obtaining the matching intention parameters from the intention parameter table according to the user intention; The node jump module is used to jump the dialogue node after completing the question and answer of the current dialogue node, and determine whether the dialogue node after the jump is the dialogue node to be filled. If so, skip the inquiry of the dialogue node to be filled and enter the next dialogue node until the dialogue process is completed.

7. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method according to any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Images