CN114117009A

CN114117009A - Method, device, equipment and medium for configuring sub-processes based on conversation robot

Info

Publication number: CN114117009A
Application number: CN202111437599.1A
Authority: CN
Inventors: 高乐
Original assignee: OneConnect Financial Technology Co Ltd Shanghai
Current assignee: OneConnect Financial Technology Co Ltd Shanghai
Priority date: 2021-11-30
Filing date: 2021-11-30
Publication date: 2022-03-01

Abstract

The invention relates to the field of data processing, and provides a sub-process configuration method, a device, equipment and a medium based on a conversation robot, wherein the method comprises the following steps: when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, and determining the distance and the topological relation between nodes in the conversation editing area according to the position coordinates; determining node attribute information corresponding to each node, and extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-flow; performing association storage on the sub-processes according to the names of the sub-processes by using the reference relation table; and calling the corresponding sub-process according to the sub-process name in the reference relation table, and configuring the sub-process according to the business process until the sub-process configuration is completed.

Description

Method, device, equipment and medium for configuring sub-processes based on conversation robot

Technical Field

The invention relates to the field of conversation robots, and provides a method, a device, equipment and a medium for identifying a sub-process configuration based on a conversation robot.

Background

With the development of artificial intelligence technology, which has entered into an unprecedented brisk development period, and also benefited from policy interest, capital push, and demand push, an intelligent conversation robot is an important direction for the development of artificial intelligence technology, and is now widely used in various fields, such as intelligent customer service, intelligent home, internet of things, and so on.

In the intelligent dialogue robot technology, a question-answer knowledge base is generally adopted in the industry, a question and an answer corresponding to the question need to be stored in the question-answer knowledge base in advance, when a user inputs the question, the intelligent dialogue robot searches the question-answer knowledge base according to the question to obtain an answer corresponding to the question and feeds the answer back to the user so as to realize man-machine dialogue, but the intelligent dialogue robot adopting the question-answer knowledge base can only support one round of dialogue. When a user carries out multiple rounds of conversations with the intelligent conversation robot, the intelligent conversation robot cannot feed back answers to the user according to context information contained in the multiple rounds of conversations, but only can independently treat each round of conversations as one independent conversation, so that when the questions input by the user in each round of conversations are not related, the intelligent conversation robot has a high probability of giving accurate answers, but when the questions input by the user in each round of conversations are related, the intelligent conversation robot cannot identify the real intention of the user and feeds back wrong or completely unrelated answers to the questions, and further the use intention and the communication efficiency of the user are reduced.

In practical application, especially the intelligent dialogue robot is configured by the process template, but with the increasing complexity of the business, the process of each dialogue robot becomes more and more complex, the process template becomes more and more huge, since the flow templates are mostly loaded and configured based on the browser, the speed of loading the flow templates is slower and slower, and, at the same time, since in most cases a robot corresponds to a business scenario, and the flow and the configuration of the business scenario are overlapped in many cases, similar functions are configured in a huge robot flow template for many times, which is time-consuming and labor-consuming, not only increases the processing burden of the conversation robot, the process of the conversation robot becomes more complex, the loading and the configuration of the process template are not facilitated, and meanwhile, the loading speed of the process template is reduced.

Disclosure of Invention

The invention provides a method, a device, equipment and a medium for configuring a sub-process based on a conversation robot, which mainly aim to send a conversation editing page request to a server according to a configuration instruction of the conversation robot by acquiring the configuration instruction of the conversation robot; receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area; when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, and determining the distance and the topological relation between nodes in the conversation editing area according to the position coordinates; and determining node attribute information corresponding to each node, extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-flow, and repeatedly calling the sub-flow to avoid repeated operation of the conversation robot when the conversation robot refers to the sub-flow, so that the processing burden of the conversation robot is reduced, the configuration of the conversation robot is simpler, the flow configuration efficiency is improved, the configuration time of the sub-flow is saved, the conversation intelligence degree of the conversation robot is promoted, and richer application scenes are met.

In order to achieve the above object, the present invention provides a sub-process configuration method based on a conversation robot, the method comprising:

acquiring a configuration instruction of a conversation robot, and sending a conversation editing page request to a server according to the configuration instruction of the conversation robot;

receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area;

when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, determining the distance between nodes in the conversation editing area according to the position coordinates, and determining the topological relation between the nodes by combining the distance between the nodes and the relative position relation;

determining node attribute information corresponding to each node, and extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-process, wherein the sub-process at least comprises a sub-process name, a sub-process category, an intention category and a sub-process description;

performing association storage on the sub-processes according to the names of the sub-processes by using a reference relation table;

and calling corresponding sub-processes according to the sub-process names in the reference relation table, and configuring the sub-processes according to the business processes until the sub-process configuration is completed.

Optionally, before the obtaining the configuration instruction of the conversation robot and sending the request of the conversation editing page to the server according to the configuration instruction of the conversation robot, the method includes:

receiving a robot information configuration instruction, wherein the information configuration instruction carries a robot identifier, and acquiring and displaying a corresponding information configuration page according to the robot identifier; and receiving an information configuration finishing instruction, acquiring basic configuration information according to the information configuration finishing instruction, and sending the basic configuration information and the robot identification to the server.

Optionally, when it is detected that a target node in a preset node area moves into a session editing area, obtaining a position coordinate of the target node in the session editing area, determining a distance between nodes in the session editing area according to the position coordinate, and determining a topological relationship between the nodes by combining the distance between the nodes and a relative position relationship, the method includes:

when detecting that a target node in a preset node area moves into a conversation editing area, acquiring the position coordinates of the target node in the conversation editing area;

analyzing and processing a connecting line between any two nodes formed by a node and a target node existing in the conversation editing area to obtain a relative position relation between any two nodes in the current conversation editing area;

and calculating the distance between any two nodes in the current conversation editing area according to the position coordinates of each node, and obtaining the topological relation between the nodes according to the distance between the nodes and the relative position relation.

Optionally, the determining node attribute information corresponding to each node, and extracting a corresponding node according to the node attribute information and the topological relation of each node in the dialog editing region to generate a sub-process includes:

extracting node configuration information of each node in the dialogue editing area, and determining node attribute information corresponding to each node according to the node configuration information of each node, wherein the node attribute information comprises a control node and a dialogue node;

forming a position relation between nodes according to a connecting line of any two nodes in the conversation editing area, and determining a topological relation between the nodes according to the position relation, wherein the type of the connecting line comprises an unintended graph line and an intended graph line;

dividing node attribute information and topological relation of each node in the dialogue editing area according to the intention of a business process to generate at least one initial sub-process;

and adjusting the node attribute information in the generated initial sub-process and the node connection line by referring to the sub-process rule to generate sub-processes which accord with the sub-process rule, wherein one sub-process at least corresponds to one version.

Optionally, before the using the reference relationship table to perform association storage on the sub-process according to the sub-process name, the method further includes:

verifying the sub-process according to a preset verification rule of the sub-process, and after the verification is qualified, performing associated storage on the sub-process according to a sub-process name and a reference relation table; wherein the preset check rule at least comprises one of the following rules:

when the current sub-process is detected to be called by the robot, the sub-process cannot be off-line; if the sub-process is to be modified, only the sub-process can be copied for modification;

and judging whether the type of a node before the end node in the current sub-process is changed, if so, and if the current sub-process is called by the robot, the sub-process cannot be used.

Optionally, the invoking a corresponding sub-process according to the sub-process name in the reference relationship table, and configuring the sub-process according to a business process includes:

extracting an intention list associated in the business process of the sub-process to be configured currently, wherein the intention list at least comprises an intention;

matching according to the intention in the intention list and the intention category corresponding to the sub-process name in the reference relation table;

when the fact that the intention type corresponding to any sub-process name is the same as or close to the intention is detected, matching is successful, and the intention association mapping between the sub-process name and the business process is determined;

and calling a sub-process of the associated mapping by using the intention of the business process, and adding the sub-process to the corresponding node for configuration by combining the position relation of each current node of the business process and the node attribute information.

Optionally, the matching according to the intention in the intention list and the intention category corresponding to the sub-flow name in the reference relationship table includes:

clustering intention categories corresponding to the sub-process names in the reference relation table by using a text clustering algorithm; obtaining a category central point of a clustering category;

calculating a distance value between the intention in the intention list and the category center point;

comparing the distance value between the intention in the intention list and the category center point with a preset distance threshold, and when the distance value between the intention in the intention list and the category center point is smaller than the preset distance threshold, indicating that the intention in the intention list is the same as or similar to the intention category corresponding to the sub-process name in the reference relation table.

In addition, to achieve the above object, the present invention provides a sub-process configuration apparatus based on a conversation robot, the apparatus including:

the acquisition module is used for acquiring a configuration instruction of the conversation robot and sending a conversation editing page request to a server according to the configuration instruction of the conversation robot;

the dialogue editing module is used for receiving and displaying a dialogue editing page returned by the server, and the dialogue editing page comprises a preset node area and a dialogue editing area;

the distance and relation determining module is used for acquiring position coordinates of a target node in a conversation editing area when the target node in a preset node area is detected to move into the conversation editing area, determining the distance between nodes in the conversation editing area according to the position coordinates, and determining the topological relation between the nodes by combining the distance between the nodes and the relative position relation;

a sub-process generation module, configured to determine node attribute information corresponding to each node, and extract a corresponding node according to the node attribute information and a topological relation of each node in the dialog editing region to generate a sub-process, where the sub-process at least includes a sub-process name, a sub-process category, an intention category, and a sub-process description;

the association storage module is used for performing association storage on the sub-processes according to the sub-process names by using the reference relation table;

and the sub-process configuration module is used for calling the corresponding sub-process according to the sub-process name in the reference relation table, and configuring the sub-process according to the business process until the sub-process configuration is completed.

Furthermore, to achieve the above object, the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the method according to any one of the above embodiments.

Furthermore, to achieve the above object, the present invention further provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps of the method according to any one of the above embodiments.

According to the sub-process configuration method, device, equipment and medium based on the conversation robot, the configuration instruction of the conversation robot is obtained, and a conversation editing page request is sent to a server according to the configuration instruction of the conversation robot; receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area; when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, and determining the distance and the topological relation between nodes in the conversation editing area according to the position coordinates; and determining node attribute information corresponding to each node, extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-flow, and repeatedly calling the sub-flow to avoid repeated operation of the conversation robot when the conversation robot refers to the sub-flow, so that the processing burden of the conversation robot is reduced, the configuration of the conversation robot is simpler, the flow configuration efficiency is improved, the configuration time of the sub-flow is saved, the conversation intelligence degree of the conversation robot is promoted, and richer application scenes are met.

Drawings

Fig. 1 is a schematic flow chart of a sub-flow configuration method based on a conversation robot according to an embodiment of the present invention;

fig. 2 is a schematic diagram illustrating a distance and relationship determination process in a sub-process configuration method based on a conversation robot according to an embodiment of the present invention;

fig. 3 is a schematic view of a sub-process generation flow in the sub-process configuration method based on the conversation robot according to an embodiment of the present invention;

FIG. 4 is a schematic view of a sub-process configuration process of the sub-process configuration method based on the conversation robot according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a sub-process configuration apparatus based on a conversation robot according to an embodiment of the present invention;

fig. 6 is a schematic structural diagram of a computer device provided in an embodiment of the present invention.

The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.

Detailed Description

It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

To facilitate understanding of the present application, the concepts related to the present application will be explained first.

In an embodiment, a flow diagram of a method for configuring a sub-flow based on a conversation robot according to an embodiment of the present invention is shown in fig. 1, and the method includes the following steps:

step S101, acquiring a configuration instruction of a conversation robot, and sending a conversation editing page request to a server according to the configuration instruction of the conversation robot;

the terminal communicates with the server through a network, receives a conversation robot configuration instruction, and sends a conversation flow editing page request to the server according to the conversation robot configuration instruction; the terminal can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers and portable wearable devices, and the server can be implemented by an independent server or a server cluster formed by a plurality of servers.

The conversation robot is a computer program that carries out a conversation via a conversation or a character, and can simulate a human conversation and pass a turing test.

Through the method, the dialog flow editing page request is generated according to the configuration instruction of the dialog robot, page editing can be performed on the dialog robot, various editing requirements of a user are further met, and the process editing capacity of the dialog robot is improved.

Step S102, receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area;

the conversation robot is a computer program that carries out conversation via conversation or characters, and can simulate human conversation and pass Turing tests. The chat robot can be used for practical purposes such as customer service, information acquisition, and speech training. The preset node is a component for carrying out conversation configuration, is preset and can be directly used. The preset nodes may include a dialogue-on-voice component, a fixed dialogue component, a fixed question and answer component, an intent dialogue component, and a scoring dialogue component, among others. The dialog editing region refers to a region for editing a dialog. Basic configuration information components are preset in the dialogue editing area. The basic configuration information component is a component generated from basic configuration information of the conversation robot, which may be sex, age, and position, etc. of the conversation robot.

Specifically, the terminal receives a conversation robot configuration instruction, sends a conversation editing page request to the server according to the conversation robot configuration instruction, the server receives the conversation editing page request, the server acquires a preset conversation editing page according to the conversation editing page request, and the conversation editing page is returned to the terminal. And the terminal receives and displays a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area, and a basic configuration information component exists in the conversation editing area.

Step S103, when it is detected that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, determining distances between nodes in the conversation editing area according to the position coordinates, and determining a topological relation between the nodes by combining the distances between the nodes and a relative position relation;

the target node refers to a component selected from each preset component, and is used for configuring a dialog in a dialog flow. The existing node refers to that a component generated according to the basic configuration information of the robot exists in the dialogue editing area or other configured components exist, and is used for determining the position of the component to be configured in the dialogue editing area.

For example, the position coordinates of the target node in the dialog editing region are acquired, the position coordinates of the existing node and the target node in the dialog editing region are acquired, the distance between any two nodes is determined by using the position coordinates between any two nodes, and the relative position relationship between any two nodes in the current dialog editing region is acquired by analyzing the connecting line between any two nodes formed by the existing node and the target node in the dialog editing region; and calculating the distance between any two nodes in the current conversation editing area according to the position coordinates of each node, and obtaining the topological relation between the nodes according to the distance between the nodes and the relative position relation.

Step S104, determining node attribute information corresponding to each node, extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-process, wherein the sub-process at least comprises a sub-process name, a sub-process category, an intention category and a sub-process description;

the topological relation includes a logical relation (topological structure relation) between the nodes, that is, the service flow between the nodes can be expressed in a directed graph manner, and the nodes reversely form the whole service flow.

By the method, the process editing mode of the conversation robot can be quickly entered, and the specific functions of each node can be clearly known through the process in the editing page and the topological relation of each node forming the process, so that the follow-up practical application is facilitated.

Step S105, the sub-process is stored in an associated manner according to the sub-process name by utilizing a reference relation table;

the reference relation table stores the sub-process parameters such as the sub-process name, the sub-process category, the intention category, the sub-process description and the like in an associated manner, for example, the reference relation table forms a mapping relation, and the subsequent calling is facilitated in a directory-like manner.

And step S106, calling corresponding sub-processes according to the sub-process names in the reference relation table, and configuring the sub-processes according to the business processes until the sub-process configuration is completed.

And calling the corresponding sub-processes by the service types or intentions of the current service processes to be configured and the sub-process names in the reference relation table, so that the sub-processes are merged into the service processes to be configured to realize the merging, and the process configuration efficiency of the conversation robot is improved.

In the embodiment, in the process of making the conversation process of the robot, a general and reusable process can be stored as a sub-process, and the sub-process is introduced in the robot editing, so that the repeated operation of the process editing is avoided, and the time for configuring the process is saved; meanwhile, sub-process parameters, functions and tasks are related and bound in the sub-processes, the sub-processes can be automatically and perfectly matched with robots for quoting the sub-processes during operation, and error prompts can be reported in editing processes at places which are not matched, so that template configuration personnel do not need to pay attention to too many operations and potential operation bugs.

The robot identifier is used for uniquely identifying the robot to be configured, the robot identifier may be a name, a numeric ID, a character string, and the like, and the information configuration page is a page for configuring basic information of the robot.

Specifically, the terminal receives a robot information configuration instruction, the information configuration instruction carries a robot identifier to be configured, the terminal analyzes the robot information configuration instruction to obtain the robot identifier, an information configuration page acquisition request is sent to the server according to the robot identifier, the server returns an information configuration page to the terminal after receiving the request, and the terminal acquires and displays the information configuration page returned by the server.

Specifically, the terminal acquires basic configuration information of the robot through an information configuration page. At the moment, the terminal receives the information configuration finishing instruction through the information configuration page, basic configuration information is obtained according to the information configuration finishing instruction, then the basic configuration information and the robot identification are sent to the server, and the server receives the basic configuration information and the robot identification and stores the basic configuration information and the robot identification in an associated mode.

In the embodiment, the basic configuration information of the robot is configured in advance through the information configuration page, and the basic configuration information and the robot mark are stored in an associated manner, so that the efficiency of the robot-to-robot interaction configuration is improved.

Optionally, when it is detected that a target node in a preset node area moves into a session editing area, obtaining a position coordinate of the target node in the session editing area, determining a distance between nodes in the session editing area according to the position coordinate, and determining a topological relationship between the nodes by combining the distance between the nodes and a relative position relationship, for example, see fig. 2, which is a schematic diagram of a distance and relationship determination process in a sub-process configuration method based on a session robot provided in an embodiment of the present invention, and is described in detail as follows:

step S201, when detecting that a target node in a preset node area moves into a conversation editing area, acquiring a position coordinate of the target node in the conversation editing area;

specifically, for example, a selected target node in the current preset node area is acquired, the target node is moved in the dialog editing area, and the position coordinate of the target node in the current dialog editing area is acquired.

For another example, point-of-interest POI data of at least two data sources of the target node are obtained, and the POI data have corresponding position coordinates; clustering position coordinates of POI data of at least two data sources to obtain at least two clustering clusters; each cluster respectively comprises at least one POI data; and respectively scoring at least two clustering clusters, and determining the position coordinates of the target nodes according to the position coordinates of the POI data in each clustering cluster corresponding to the scoring result.

Step S202, analyzing and processing a connecting line between any two nodes formed by a node and a target node existing in a conversation editing area to obtain a relative position relation between any two nodes in the current conversation editing area;

specifically, a connection line between any two nodes is analyzed (for example, by using a regular expression rule) to obtain a relative position relationship between any two nodes in the current dialog editing region, for example, a connection relationship, a directed relationship, and the like between the nodes.

Step S203, calculating the distance between any two nodes in the current dialogue editing area according to the position coordinates of each node, and obtaining the topological relation between the nodes according to the distance between the nodes and the relative position relation.

Specifically, after the position coordinates of any two nodes are obtained, the distance between the nodes can be determined by using the coordinate distance (such as the pythagorean theorem), and the topological relation between the nodes is obtained according to the distance between the nodes and the relative position relation.

In this embodiment, the distance and the relative position relationship between the nodes in the current robot conversation process can be obtained to obtain the topological relationship between the nodes, which is beneficial to subsequent segmentation or extraction of one or more sub-processes.

Optionally, the node attribute information corresponding to each node is determined, and a corresponding node is extracted according to the node attribute information and the topological relation of each node in the dialogue editing area to generate a sub-flow, which is shown in detail in fig. 3, and is a schematic diagram of a sub-flow generation flow in a sub-flow configuration method based on a dialogue robot provided in an embodiment of the present invention, and detailed below:

step S301, extracting node configuration information of each node in the dialogue editing area, and determining node attribute information corresponding to each node according to the node configuration information of each node, wherein the node attribute information comprises a control node and a dialogue node;

the method comprises the steps that each node has initial configuration information during initial configuration, the node configuration information of each node in a dialogue editing area is extracted, if the node configuration information is not changed, the initial configuration information is used as the configuration information of the node which is not changed, and meanwhile, if the node configuration information is changed, the configuration information which is changed by the node is called.

Step S302, according to the position relation between nodes formed by the connection lines of any two nodes in the dialogue editing area, determining the topological relation between the nodes according to the position relation, wherein the types of the connection lines comprise non-intention lines and intention lines;

in the dialogue editing area, connecting lines of all nodes form a position relation among the nodes, a topological relation among the nodes is determined according to the position relation, and corresponding non-intention lines or intention lines are selected for the control nodes or the dialogue nodes according to the types of the connecting lines.

Step S303, dividing node attribute information and topological relation of each node in the dialogue editing area according to intention of a business process to generate at least one initial sub-process;

wherein, as the dialog editing area may have a plurality of intentions, a plurality of sub-flows can be divided according to the types of the intentions.

Step S304, referring to the sub-process rules, adjusting the node attribute information and the node connection line in the generated initial sub-process, and generating sub-processes according with the sub-process rules, wherein one sub-process at least corresponds to one version.

The sub-process rule may be set manually, for example, the type and the connection type of a node before an end node of a certain sub-process may be set, which is not described herein again.

For example, by adding the sub-processes, the method further includes:

acquiring an optional sub-process list by using the newly added sub-process drawing node; and determining a certain sub-process in the sub-process table, storing the service ID of the sub-process into the sub-process by using the front end, and storing the type of the sub-process node.

The sub-process list comprises a conversation robot ID, a phone-to-phone robot name, a robot issuing state and an operation process; the sub-processes also comprise a voice sub-process and a text sub-process, so that the voice robot can only refer to the voice sub-process, and the text robot can only refer to the text sub-process.

By the method, in the process of making the conversation process of the conversation robot, the universal and reusable process is stored as the sub-process, and the sub-process is introduced in the robot editing, so that the repeated operation of the process editing is avoided, and the time for configuring the process is saved.

when the current sub-process is detected to be called by the robot, the sub-process cannot be off-line; if the sub-process is to be modified, the sub-process can be copied only and modified in a new template mode;

and judging whether the type of a node before the end node in the current sub-process is changed, if so, and if the current sub-process is called by the robot, the sub-process cannot be used (namely, the sub-process is issued).

For example, when the relation between each conversation robot and the sub-process in the reference relation table is used, the reference relation table of the conversation robot and the sub-process is obtained; when a saving and issuing instruction is received, whether the parameter set of the saving and issuing interface and the parameter set of the sub-process referenced by all the conversation robots have parameters with the same name and different types or not is judged, if not, the verification is successful, the issuing operation and the saving operation can be carried out, and if yes, the verification is failed, and the operation is not carried out.

Specifically, the intentions, the function functions (each sub-process node includes a function, each function has a function name, and the parameters of each function include a processing class and an entity variable) of the sub-processes and the tasks are configured in the templates corresponding to the sub-processes without checking; by calling the parameters, functions, tasks and other modes in the sub-process, the perfect fit of the robot process configuration is realized, the error prompt can be reported in the editing process at the non-fit place, meanwhile, the template configuration personnel do not need to pay much attention, and many operations and potential running bugs are simplified.

When the front-end graph is detected to be converted into a back-end graph, replacing the conversion logic of the new sub-process, wherein a sub-process node object is newly added in the back-end graph, and the sub-process node stores a sub-process service ID;

and when the conversation robot storage and release interface is received to finish the newly added check, determining all the quoted sub-processes as the release state, and storing and releasing the conversation robot.

In other embodiments, if the sub-process is referred to by the issued robot, the sub-process cannot be directly offline, and if the sub-process is modified, the sub-process can be copied, and the new version of the sub-process is modified and issued to replace the originally issued sub-process.

When the sub-process is issued, the cache of the robot which refers to the sub-process needs to be cleaned.

Through the mode, the conversation robot storage and release interface is adjusted and verified, the conversation robot storage and release interface can be stored and released only when the verified sub-process is met, the accuracy of the sub-process can be ensured, and the phenomenon that the sub-process is randomly introduced subsequently is avoided.

Optionally, the corresponding sub-process is called according to the sub-process name in the reference relationship table, and the sub-process is configured according to the business process, which is detailed in fig. 4, and is a schematic diagram of the sub-process configuration process in the sub-process configuration method based on the conversation robot provided in an embodiment of the present invention, which is detailed as follows:

step S401, extracting an intention list associated in the business process of the sub-process to be configured currently, wherein the intention list at least comprises an intention;

the intention can be generated according to the keywords in the current conversation process, so as to form an intention list, for example, the pre-trained Bert model is adopted to perform intention recognition on the keywords in the current conversation process, and an intention recognition result is obtained.

Step S402, matching according to the intention in the intention list and the intention category corresponding to the sub-process name in the reference relation table;

and matching the intentions by means of similarity or identity of the intentions and intention categories corresponding to the sub-process names in the reference relation table.

Step S403, when detecting that the intention type corresponding to any sub-process name is the same as or close to the intention, matching is successful, and determining the intention association mapping between the sub-process name and the business process;

whether the intention type corresponding to any sub-process name is the same as or similar to the intention or not is judged through similarity, and details are not repeated herein.

Step S404, utilizing the intention of the business process to call the sub-process of the associated mapping, and adding the sub-process to the corresponding node for configuration by combining the position relation of each current node of the business process and the node attribute information.

The sub-processes are stored in a template form, each sub-process template is stored in a directory form in an associated mode through a reference relation table, the corresponding sub-process template can be called according to the current to-be-configured business process, the number of the called sub-processes is not limited, and calling is carried out according to the requirements of the business process.

for example, dividing intentions corresponding to the sub-process names in the reference relation table into a plurality of text categories by adopting a text clustering algorithm; after the label library is constructed, the labels in the label library can be clustered into a plurality of text categories by adopting a text clustering algorithm (for example, tf-idf + kmeans), and the clustering of the intentions is completed.

the distance values between the graph and the class center point are determined, for example, by one or several of euclidean distance, cosine similarity, manhattan distance, minkoch distance, jaccard similarity coefficient, pearson correlation coefficient, hamming distance, or BM25 correlation score.

For example, by determining whether the intention in the intention list of the sub-flow to be configured currently is the same as or similar to the intention category corresponding to the sub-flow name in the reference relation table, the intention whose distance value meets the requirement of the preset distance threshold is regarded as a successful matching.

In the embodiment, clustering is performed according to the intention category of the node to generate a clustering result, and the adopted unidentified intention text is configured in the intention similarity operation, so that the robot intention recognition capability is efficiently optimized in a targeted manner.

Optionally, in other embodiments, each sub-process is called by using a sub-process name in the reference relationship table, and each sub-process is merged according to a service process to obtain a merged flow template of the telephone robot, where details of a manner in which each sub-process is merged according to a service flow are as follows:

and when the type of the corresponding node in the analyzed flow template is detected, calling the sub-flow stored by the service ID in the back-end server according to the sub-flow service ID corresponding to the current node, analyzing, and combining all the nodes in the flow template of the conversation robot according to a preset rule.

All father nodes corresponding to the end nodes in the sub-process of the conversation robot are changed to point to the lower nodes pointed by the sub-process nodes in the conversation robot;

changing a father node corresponding to a child process node of the conversation robot into a first node of a child process;

and if the starting node of the conversation robot is just the sub-process node, setting the starting node of the conversation robot as the starting node in the sub-process.

Specifically, the attribute information of the node at least includes the type of the node, the sub-process stored in the back-end server is called for analyzing the service ID according to the sub-process service ID corresponding to the current node, and the sub-processes are merged in the main process according to the preset rule in combination with the current service process, wherein the merging mode includes but is not less than superposition and fusion. For example, the sub-processes are split according to the starting node, the intermediate node and the ending node, and the process template fusion among different sub-processes is realized according to the service logic among all the nodes.

For example, in the exit part of the editing sub-process, and under the robot that refers to the sub-process, many restrictions and prompts are made on the connecting lines and other corresponding rules in the editing interface, so that the template configurator does not pay much attention to the complex logic, and the program can help him to adjust or prompt.

In some embodiments, the adjusting by DM (process management) to adjust the processing flow of the new sub-process node further includes:

1) a sub-process, in which a line connecting to the end node is an unintended line;

2) one node before the end node cannot be used as a logic node;

3) one node before the end node can be used as a dialect node or a control node;

4) when the sub-process is issued, the sub-process before the change is checked, whether the type of the node before the node is changed or not is judged, if yes, the sub-process cannot be issued because the current sub-process is quoted by the inventor.

Specifically, the rules and complex logic in the original sub-process can be weakened by setting the limiting conditions through the connecting lines in the sub-process editing.

Through the mode, on the basis of the original main process, after the specific speech operation node is triggered, other sub-processes (speech operation branches) are called until the conversation is finished, the original main process is not returned, the intelligent degree of the robot conversation is improved by configuring the speech operation sub-processes, and richer scene requirements are met.

It should be noted that the sub-process management table has the following structure:

and (4) a sub-flow chart:

CREATE TABLE public.aicsp_robot_subflow(

id serial NOT NULL, primary key id

name varchar (60) NULL, sub-process name

business _ Id int4 NOT NULL, sub-process business Id, business Id of concept and robot (robot _ Id)

status int2 NULL DEFAULT 0, -completion status 0: non-release 1: issue a publication

description varchar (300) NULL, a sub-flow description

version _ id varchar (30) NULL, -sub-flow version number

version _ desc varchar (1000) NULL, sub-process version description

robot _ param jsonb NULL, I/O parameter of subroutine call

robot _ info jsonb NULL, -front end map

front _ graph jsonb NULL, -Back end graph

nlp _ lab jsonb NULL, the value corresponding to the intntLibrary inside is the intention library

category varchar (30) NULL, sub-process category, classified as CALL and IM

apple varchar (100) NULL, -tenant id

created _ user varchar (60) NULL, -create user

created _ time _ NULL, -creation time

updated user varchar (60) NULL, -update user

updated time parameter NULL

)；

Many-to-many association table of sub-processes and robots:

CREATE TABLE public.aicsp_robot_subflow_rel(

id serial NOT NULL, primary key id

robot _ id int4 NOT NULL, -robot primary key id

subflow _ id int4 NOT NULL, business id of the sub-process (business _ id)

created _ user varchar (60) NULL, -create user

created _ time _ NULL, -creation time

updated user varchar (60) NULL, -update user

updated time parameter NULL

The type of the line connected with the sub-process node is determined by the type of the previous node of the internal end node of the sub-process;

if the node is a dialect node, the sub-process node connecting line is an ideogram;

if the node is the control node, the sub-process node connecting line is an unintended line;

in the sub-process nodes, when the sub-process is replaced, if the sub-process nodes have connecting lines and the types of the previous nodes of the internal end nodes of the new sub-process and the previous nodes of the internal end nodes of the old sub-process are different, a prompt pops up.

If the control node is a dialect node, the type of the replacement connecting line is an intention line;

if the operating node is greater than the control node, only one connecting line is provided, and the type of the replacing connecting line is an unintended graph line; and if a plurality of connecting lines exist, deleting all the connecting lines which are intented to be connected.

The embodiment provides a sub-process configuration method based on a conversation robot, which comprises the steps of sending a conversation editing page request to a server according to a configuration instruction of the conversation robot by acquiring the configuration instruction of the conversation robot; receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area; when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, and determining the distance and the topological relation between nodes in the conversation editing area according to the position coordinates; and determining node attribute information corresponding to each node, extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-flow, and repeatedly calling the sub-flow to avoid repeated operation of the conversation robot when the conversation robot refers to the sub-flow, so that the processing burden of the conversation robot is reduced, the configuration of the conversation robot is simpler, the flow configuration efficiency is improved, the configuration time of the sub-flow is saved, the conversation intelligence degree of the conversation robot is promoted, and richer application scenes are met.

In one embodiment, the present invention further provides a conversation robot-based sub-process configuration apparatus 500, referring to fig. 5, the apparatus comprising:

an obtaining module 501, configured to obtain a configuration instruction of a conversation robot, and send a conversation editing page request to a server according to the configuration instruction of the conversation robot;

the session editing module 502 is configured to receive and display a session editing page returned by the server, where the session editing page includes a preset node region and a session editing region;

a distance and relationship determining module 503, configured to, when it is detected that a target node in a preset node area moves into a session editing area, obtain a position coordinate of the target node in the session editing area, determine a distance between nodes in the session editing area according to the position coordinate, and determine a topological relationship between nodes by combining the distance between the nodes and a relative position relationship;

a sub-process generating module 504, configured to determine node attribute information corresponding to each node, and extract a corresponding node according to the node attribute information and the topological relation of each node in the dialog editing region to generate a sub-process, where the sub-process at least includes a sub-process name, a sub-process category, an intention category, and a sub-process description;

an association storage module 505, configured to perform association storage on the sub-processes according to sub-process names by using a reference relationship table;

and a sub-process configuration module 506, configured to call a corresponding sub-process according to the sub-process name in the reference relationship table, and configure the sub-process according to the business process until the sub-process configuration is completed.

Optionally, in some embodiments, before the obtaining module 501, the method further includes: configuration information module for

Based on the above embodiment, the distance and relationship determining module 503 further includes: the method comprises the steps that when a target node in a preset node area is detected to move into a conversation editing area, the position coordinates of the target node in the conversation editing area are obtained; analyzing and processing a connecting line between any two nodes formed by a node and a target node existing in the conversation editing area to obtain a relative position relation between any two nodes in the current conversation editing area; and calculating the distance between any two nodes in the current conversation editing area according to the position coordinates of each node, and obtaining the topological relation between the nodes according to the distance between the nodes and the relative position relation.

Based on the above embodiment, the sub-process generating module 504 further includes: extracting node configuration information of each node in the dialogue editing area, and determining node attribute information corresponding to each node according to the node configuration information of each node, wherein the node attribute information comprises a control node and a dialogue node; forming a position relation between nodes according to a connecting line of any two nodes in the conversation editing area, and determining a topological relation between the nodes according to the position relation, wherein the type of the connecting line comprises an unintended graph line and an intended graph line; dividing node attribute information and topological relation of each node in the dialogue editing area according to the intention of a business process to generate at least one initial sub-process; and adjusting the node attribute information in the generated initial sub-process and the node connection line by referring to the sub-process rule to generate sub-processes which accord with the sub-process rule, wherein one sub-process at least corresponds to one version.

Optionally, in some embodiments, before the associating and storing module 505, further include: the verification module is used for verifying the sub-process according to a preset verification rule of the sub-process, and storing the sub-process in an associated manner according to the sub-process name and the reference relation table after the sub-process is verified to be qualified; wherein the preset check rule at least comprises one of the following rules:

Optionally, based on the foregoing embodiment, the sub-process configuration module 506 further includes: extracting an intention list associated in the business process of the sub-process to be configured currently, wherein the intention list at least comprises an intention; matching according to the intention in the intention list and the intention category corresponding to the sub-process name in the reference relation table; when the fact that the intention type corresponding to any sub-process name is the same as or close to the intention is detected, matching is successful, and the intention association mapping between the sub-process name and the business process is determined; and calling a sub-process of the associated mapping by using the intention of the business process, and adding the sub-process to the corresponding node for configuration by combining the position relation of each current node of the business process and the node attribute information.

Optionally, in the above embodiment, the matching according to the intention in the intention list and the intention category corresponding to the sub-flow name in the reference relationship table includes:

The embodiment provides a sub-process configuration device based on a conversation robot, which sends a conversation editing page request to a server according to a configuration instruction of the conversation robot by acquiring the configuration instruction of the conversation robot; receiving and displaying a conversation editing page returned by the server, wherein the conversation editing page comprises a preset node area and a conversation editing area; when detecting that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, and determining the distance and the topological relation between nodes in the conversation editing area according to the position coordinates; and determining node attribute information corresponding to each node, extracting corresponding nodes according to the node attribute information and the topological relation of each node in the conversation editing area to generate a sub-flow, and repeatedly calling the sub-flow to avoid repeated operation of the conversation robot when the conversation robot refers to the sub-flow, so that the processing burden of the conversation robot is reduced, the configuration of the conversation robot is simpler, the flow configuration efficiency is improved, the configuration time of the sub-flow is saved, the conversation intelligence degree of the conversation robot is promoted, and richer application scenes are met.

It should be understood that, the device for configuring a sub-process based on a conversation robot is substantially provided with a plurality of modules for executing the method for configuring a sub-process based on a conversation robot in any of the embodiments, and specific functions and technical effects are only required to refer to the embodiments, which are not described herein again.

In an embodiment, referring to fig. 6, the embodiment further provides a computer device 600, which includes a memory 601, a processor 602, and a computer program stored on the memory and executable on the processor, and when the processor 602 executes the computer program, the steps of the method according to any one of the above embodiments are implemented.

In an embodiment, a computer-readable storage medium is also provided, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any of the above embodiments.

The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, apparatus, article, or method that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, apparatus, article, or method. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, apparatus, article, or method that includes the element.

The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments. Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, or a network device) to execute the method according to the embodiments of the present invention.

The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims

1. A method for configuring a sub-process based on a conversation robot, the method comprising:

2. The sub-process configuration method based on the conversation robot as claimed in claim 1, wherein the step of obtaining the configuration command of the conversation robot before sending the request of the conversation editing page to the server according to the configuration command of the conversation robot comprises:

3. The method for configuring sub-processes based on conversation robots according to claim 1, wherein when it is detected that a target node in a preset node area moves into a conversation editing area, acquiring position coordinates of the target node in the conversation editing area, determining distances between nodes in the conversation editing area according to the position coordinates, and determining topological relations between nodes by combining the distances between the nodes and the relative position relations, comprises:

4. The method for configuring sub-processes based on conversation robots according to any one of claims 1 to 3, wherein the determining of the node attribute information corresponding to each node and the extracting of the corresponding node according to the node attribute information and the topological relation of each node in the conversation editing area to generate the sub-processes comprises:

5. The method of claim 4, wherein before storing the sub-processes in association with each other according to their names by using a reference relationship table, the method further comprises:

6. The conversation robot-based sub-process configuration method according to any one of claims 1 to 3, wherein the step of calling the corresponding sub-process according to the sub-process name in the reference relationship table and configuring the sub-process according to the business process comprises:

7. The conversation robot-based sub-process configuration method according to claim 6, wherein the matching according to the intentions in the intention list according to the intention categories corresponding to the sub-process names in the reference relationship table comprises:

8. A conversation robot-based sub-process configuration apparatus, the apparatus comprising:

9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the steps of the method of any of claims 1 to 7 are implemented by the processor when executing the computer program.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.