Process approval intelligent interaction method, system and device based on large model and medium

Abstract

The invention provides a process approval intelligent interaction method, system and device based on a large model and a medium, and belongs to the technical field of office automation. The method comprises the steps that a user behavior rule base is created, the mapping relation between query natural languages and task query variables is stored, and user behavior habits are recorded; establishing an association relationship between the task table and the task information source; responding to an approval query request of a user, converting the approval query request into a query natural language through the large model, and generating a task query variable according to the mapping relationship; executing a task query variable based on the association relationship, querying all tasks meeting conditions, and displaying the tasks to the user; responding to an examination and approval request of a user, converting the examination and approval request into an examination and approval natural language through the large model, analyzing, determining an examination and approval action and a target task, and executing the examination and approval action; and recording the user approval behavior process to the user behavior rule. According to the invention, natural language processing and the user behavior rule base are combined, so that an efficient and intelligent personalized approval process is realized.

Description

Technical Field

[0001] The present invention belongs to the field of office automation technology, and specifically relates to a large model-based process approval intelligent interaction method, system, equipment and medium. Background Art

[0002] With the development of digital office, process approval is an important part of daily business operations, and its efficiency and interactive experience directly affect the operation of the enterprise. User approval operations in traditional ERP require cumbersome steps: first, you need to navigate to the task list in the APP; then filter out the tasks to be processed from the long task list; finally, perform approval operations for each specific task one by one. This process relies on a large number of mobile phone screen clicks and interactive operations, which is not only complicated to operate, but also time-consuming, seriously affecting work efficiency.

[0003] With the development of artificial intelligence technology, intelligent question and answer as an information interaction method of natural language processing technology has gradually become popular. In the specific scenario of process approval, it has become a trend to use the language understanding and generation capabilities of large models such as GPT and BERT to provide users with efficient and convenient user experience. However, there are still many problems in applying large models to process approval. First, the existing human-computer interaction model used in task approval is relatively traditional and rigid, lacking intelligent interaction methods. Users have to adapt to the preset operation process of the system, and cannot flexibly adjust it according to their own usage habits or special needs. This interaction method not only limits the user's freedom of operation, but also reduces the efficiency and satisfaction of approval work. On the other hand, when using large models to process task queries and approvals, the capture and utilization of individual user behavior habits are insufficient, making it difficult to fully meet the personalized needs of users, and unable to maximize the user experience, and the approval efficiency has not been maximized. Summary of the invention

[0004] In a first aspect, an embodiment of the present application provides a process approval intelligent interaction method based on a large model, comprising the following steps: S1. Create a user behavior rule library to store the mapping relationship between query natural language and task query variables, and record user behavior habits; S2. Determine the table related to task approval as the task information source, and establish the association relationship between the task table and the task information source; S3. respond to the user's approval query request, and convert the approval query request into query natural language through the large model, and generate task query variables according to the mapping relationship between the query natural language and the task query variables; S4. executing the task query variable based on the association between the task table and the task information source, querying all tasks that meet the conditions and displaying them to the user; S5. Based on the task queue, respond to the user's approval request, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; S6. Record the user's approval behavior process into the user behavior rule library.

[0005] Furthermore, the specific steps of step S1 are as follows: S11. Define the task information fields on the task table; S12. Count the tasks in the task table according to the application scenario, and count the natural language queries in each scenario; S13. Extract keywords from the query natural language of each type of scenario in combination with the task information field defined in the task table to construct a task query variable; S14. Establish a mapping relationship between the query natural language and the task query variable, and save it in the user behavior rule library; S15. Analyze the user's historical query behavior to identify the user's behavior habits, and save them to the user behavior rule library.

[0006] Furthermore, the specific steps of step S2 are as follows: S21 task information corresponding to the task information field in the task table is stored; S22. Determine the table related to task approval as the task information source; S23. Add the fields related to task approval in each task information source to the task table, and establish an association relationship between the task table and each task information source.

[0007] Furthermore, the specific steps of step S3 are as follows: S31. Identify the user's approval query request and respond to it, and provide it to the natural language processing module of the large model; S32. The natural language processing module converts the user's approval query request into natural language form, and obtains the task query variable and the corresponding variable value according to the mapping relationship between the natural language and the task query variable in the user behavior rule library.

[0008] Furthermore, the specific steps of step S4 are as follows: S41. Determine the required task information source according to the task query variable; S42. Query the task table according to the variable value of the task query variable and the required task information source, determine all tasks that meet the conditions, generate a task queue, and display it to the user; S43. Identify whether the user's input is a re-approval query request or an approval request; If it is an approval request, go to step S5; When the query request is to be reviewed again, the process goes to step S44; S44. Integrate the re-approval query request with the previous approval query request and use it as the approval query request, and return to step S32.

[0009] Furthermore, the specific steps of step S5 are as follows: S51. Based on the task queue, identify the user's approval request and respond to it, and provide it to the natural language processing module of the large model; S52. The natural language processing module converts the user's approval request into natural language form and extracts the approval action and target task; S53. Generate confirmation prompts based on the approval action and target task, and return to the user; S54 responds to the user's confirmation of the confirmation prompt word and checks whether the approval conditions are met; If yes, go to step S57; If not, proceed to step S55; S55. Generate approval prompt suggestions based on approval conditions and user behavior habits in the user behavior rule library; S56. respond to the user's feedback on the approval prompt suggestion and check whether the approval conditions are met; If yes, go to step S57; If not, return to step S55; S57. Execute the approval action on the target task in combination with the approval conditions to complete the approval operation.

[0010] Furthermore, the specific steps of step S6 are as follows: S61. Analyze multiple approval query requests in the user's current approval behavior; S62. Identify the task query variables that have been increased or decreased in the subsequent approval query request compared to the previous approval query request, and mark the query using the condition type, and then add it to the user behavior habits of the user behavior rule library; S63. Count the operations on the approval conditions before the user performs the approval action in this approval behavior, and add the statistical results to the user behavior habits in the user behavior rule library.

[0011] In a second aspect, the embodiment of the present application further provides a process approval intelligent interactive system based on a large model, including: A user behavior rule base building module is used to create a user behavior rule base, store the mapping relationship between query natural language and task query variables, and record user behavior habits; An information source association relationship establishment module is used to determine a table related to task approval as a task information source and establish an association relationship between the task table and the task information source; The task query variable generation module is used to respond to the user's approval query request, convert the approval query request into query natural language through the big model, and generate task query variables according to the mapping relationship between the query natural language and the task query variable; The approval task query module is used to execute task query variables based on the association between the task table and the task information source, query all tasks that meet the conditions, generate a task queue, and display it to the user; The approval task execution module is used to respond to the user's approval request based on the task queue, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; Add a module to the user behavior rule library to record the user's approval behavior process in the user behavior rule library.

[0012] In a third aspect, an embodiment of the present application further provides an electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein when the processor executes the program, the steps of the intelligent interactive method for process approval based on a large model as described in the first aspect are implemented.

[0013] In a fourth aspect, an embodiment of the present application further provides a storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of the intelligent interactive method for process approval based on a large model as described in the first aspect are implemented.

[0014] It can be seen from the above technical solutions that the present invention has the following advantages: In the intelligent interactive method, system, device and medium for process approval based on big model provided by the present application, the user's approval query and approval request are converted into natural language processing through the big model, which changes the traditional cumbersome manual operation mode and saves the user's time in searching and approving tasks, so that the user does not need to manually screen in a complex task list, but can directly describe the requirements through natural language to quickly locate and process related tasks, thereby improving the efficiency of process approval; the user behavior rule library created can record user behavior habits, and combined with the powerful capabilities of the big model, provide users with personalized approval prompts and suggestions, and in the approval process, give approval prompts and suggestions based on the user's historical behavior and current task characteristics, assist user decision-making, and improve user experience; by establishing an association between the task table and related information sources, the query data can be integrated more efficiently, so that in the task query and approval process, data can be quickly obtained from multiple information sources, providing a data basis for subsequent data analysis and process optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solution of the present invention, the accompanying drawings required for use in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0016] Figure 1 The figure is a flow chart of the intelligent interactive method for process approval based on a large model of the present invention.

[0017] Figure 2 It is a schematic diagram of the process approval intelligent interactive system based on a large model of the present invention. DETAILED DESCRIPTION

[0018] In the following, the specific steps of the intelligent interactive method for process approval based on a large model will be described in detail, and various embodiments of the present disclosure will be described more comprehensively. The present disclosure may have various embodiments, and adjustments and changes may be made therein. However, it should be understood that there is no intention to limit the various embodiments of the present disclosure to the specific embodiments disclosed herein, but rather the present disclosure should be understood to cover all adjustments, equivalents and / or alternatives that fall within the spirit and scope of the various embodiments of the present disclosure.

[0019] For example, with the development of digital office, process approval is the core link of enterprise operation, and its efficiency and interactive experience have a direct impact on the overall operation efficiency of the enterprise. In the traditional ERP system, users often have to go through cumbersome steps to perform approval operations: first enter the APP application and find the task list; then filter out the tasks to be processed from the numerous tasks; and finally approve each task one by one. This process relies heavily on the click operation of the mobile phone screen, which is not only complicated but also time-consuming, greatly reducing work efficiency.

[0020] With the development of artificial intelligence technology, intelligent question and answer as an interactive method of natural language processing technology is becoming increasingly popular. In the specific field of process approval, using the language understanding and generation capabilities of large models such as GPT and BERT to provide users with a more efficient and convenient user experience has become a new development trend. However, the application of large models to process approval still faces many challenges. On the one hand, the human-computer interaction mode adopted by the existing task approval system is relatively traditional and rigid, and lacks intelligent interaction means. Users can only operate according to the preset process of the system and cannot flexibly adjust according to their own habits or needs. This not only limits the user's freedom of operation, but also reduces the approval efficiency and user satisfaction. On the other hand, when using large models to process task queries and approvals, the system's capture and utilization of individual user behavior habits is not sufficient, making it difficult to meet the personalized needs of users and unable to improve the user experience and approval efficiency to the best state.

[0021] In response to the above problems, this embodiment provides a large-model-based intelligent interactive method for process approval. Through the combination of natural language processing and user behavior rule library, intelligent approval query and approval operation are realized, which significantly improves user experience and approval efficiency.

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1 The figure is a flowchart of a process approval intelligent interaction method based on a large model in a specific embodiment, and the method includes the following steps: S1. Create a user behavior rule library to store the mapping relationship between query natural language and task query variables, and record user behavior habits; It should be noted that by storing the mapping relationship between query natural language and task query variables, the approval system can understand the user's query natural language query intention, establish a connection between the user's daily query natural language and the task query variables within the system, and improve the accuracy of the query; by recording user behavior habits, the query results and approval suggestions are optimized according to the user's historical operation habits, thereby improving the user experience and providing users with personalized services; S2. Determine the table related to task approval as the task information source, and establish the association relationship between the task table and the task information source; It should be noted that by establishing the association between the task table and the task information source, accurate data can be obtained from multiple relevant information sources when performing task query and approval during the approval process, thus avoiding the problem of missing or inconsistent data, improving the efficiency of data query, ensuring the reliability of data during task processing, and providing a data basis for subsequent approval operations; S3. respond to the user's approval query request, and convert the approval query request into query natural language through the large model, and generate task query variables according to the mapping relationship between the query natural language and the task query variables; It should be noted that the natural language processing capability of the big model can transform the user's query request into a task query variable that the approval system can understand, thus achieving the connection between user needs and system processing; compared with the traditional query method, the flexibility and accuracy of the query are improved, and users can express their query needs in a more flexible way, and the big model can quickly and accurately generate the corresponding task query variables; S4. executing the task query variable based on the association between the task table and the task information source, querying all tasks that meet the conditions, generating a task queue, and displaying it to the user; It should be noted that, according to the generated task query variable, a quick query can be performed in the task table and information source with established associations, and tasks that meet user needs can be screened out and displayed to the user, ensuring that the user can quickly obtain accurate task information and improving the efficiency of task query; S5. Based on the task queue, respond to the user's approval request, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; It should be noted that the big model is used to parse the user's approval request to accurately identify the approval action and target task; by checking the approval conditions and generating approval prompts based on the user's behavior habits, it helps the user make reasonable approval decisions, and executes the approval action after the approval conditions are met, ensuring that the approval process meets the requirements; S6. Record the user's approval behavior process into the user behavior rule library; It should be noted that by recording the user's approval behavior process in approval query requests, approval actions, and approval condition processing, the approval system can analyze the user's behavior patterns, improve the user behavior rule library, and provide accurate and personalized services for future approval operations.

[0024] This embodiment realizes the intelligence and personalization of the approval process by creating a user behavior rule library for the task center, storing the mapping relationship between query natural language and task query variables, and recording user behavior habits. The accuracy of approval data is ensured by establishing an association between the task table and the task information source; the efficiency and accuracy of approval are improved by responding to user approval queries and approval requests and using a large model for natural language processing and parsing; and the user's approval behavior process is recorded in the user behavior rule library, which provides a basis for the optimization of the user behavior rule library.

[0025] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process in this embodiment, another intelligent interactive method for process approval based on a large model is provided, and the method includes the following steps: S1. Create a user behavior rule library to store the mapping relationship between query natural language and task query variables, and record user behavior habits; the specific steps of step S1 are as follows: S11. Define the task information fields on the task table; For example, the task information fields may include the initiator, task type (such as expense report, leave application, project approval), task creation time, task approval completion time, task urgency, task status (such as pending approval, approved, rejected), whether it is overdue, task form type, and the initiator's department; S12. Count the tasks in the task table according to the application scenario, and count the natural language queries in each scenario; In this embodiment, the task-related intelligent questions and answers are divided into two categories: one is task query, which may be a task that has been approved in the past period of time, or a task that is about to expire and needs to be approved as soon as possible; the other is task approval, which locates the task for approval based on the result of the task query, mainly involving approval, rejection, transfer, and retrieval; the approval of the task depends on the query of the task, and after querying the task, the relevant approval operations are performed according to the needs of the user; S13. Extract keywords from the query natural language of each type of scenario in combination with the task information field defined in the task table to construct a task query variable; For example, by querying the natural language "find the official documents submitted by XXX yesterday", the task query variables and variable values ​​shown in Table 1 can be obtained; Table 1

[0026] S14. Establish a mapping relationship between the query natural language and the task query variable, and save it in the user behavior rule library; For example, the mapping relationship between the query natural language and the task query variable is shown in Table 2: Table 2

[0027] S15. Analyze the user's historical query behavior to identify the user's behavior habits and save them to the user behavior rule library; For example, user A often searches for "expense reimbursement forms submitted by the Finance Department this week", and uses "this week" and "expense reimbursement forms of the Finance Department" as common search conditions; user B often adds "urgency" as a filter condition after searching, and marks "urgency" as a common condition; It should be noted that, through the definition of task information fields in the task table, natural language statistics, keyword extraction, and mapping relationship establishment, the approval system can accurately understand the user's natural language query needs and convert them into effective task query variables; at the same time, it analyzes the user's historical query behavior to identify user behavior habits, provides data support for personalized services, and improves the approval system's accurate response to user needs; S2. Determine the table related to task approval as the task information source, and establish the association between the task table and the task information source; the specific steps of step S2 are as follows: S21. Storing the task information corresponding to the task information field in the task table; S22. Determine the table related to task approval as the task information source; Exemplarily, a task table is established to store basic information of the task; Determine the user table and department table as the task information source. Use the user table to store the user's basic information, such as name, department, and employment status. Use the department table to store department information, such as department name and department head. S23. Add the fields related to task approval in each task information source to the task table, and establish an association relationship between the task table and each task information source; For example, the "Originator ID" field in the task table is associated with the "User ID" in the user table, the "Approver ID" field in the task table is associated with the "User ID" in the user table, and the "Department ID" field in the user table is associated with the "Department ID" in the department table; It should be noted that by establishing a task table to store task information, determining the relevant task information source, and adding related fields, it is ensured that the required information can be quickly and accurately obtained from multiple data sources during the task query and approval process, thereby improving data query efficiency and providing a data basis for subsequent task processing; S3. Respond to the user's approval query request, and convert the approval query request into query natural language through the large model, and generate task query variables according to the mapping relationship between the query natural language and the task query variable; the specific steps of step S3 are as follows: S31. Identify the user's approval query request and respond to it, and provide it to the natural language processing module of the large model; S32. The natural language processing module converts the user's approval query request into natural language form, and obtains the task query variable and the corresponding variable value according to the mapping relationship between the natural language and the task query variable in the user behavior rule library; For example, the user inputs "find the expense reports submitted by the Finance Department this week" through voice or text; the approval system obtains the task query variable through the natural language processing module: Sponsor Department: Finance Department Task type: Expense reimbursement Task creation time: this week It should be noted that by identifying the user's approval query request, the natural language processing module of the big model is used to convert the request into natural language, and the task query variables and corresponding variable values ​​are obtained according to the established mapping relationship, so as to realize the rapid connection between the user's query demand and the approval system data processing, and improve the accuracy and efficiency of the task query; S4. Based on the relationship between the task table and the task information source, the task query variable is executed to query all tasks that meet the conditions, generate a task queue, and display it to the user; the specific steps of step S4 are as follows: S41. Determine the required task information source according to the task query variable; S42. Query the task table according to the variable value of the task query variable and the required task information source, determine all tasks that meet the conditions, generate a task queue, and display it to the user; For example, the approval system first finds all users in the finance department from the user table based on the extracted task query variable, then filters out the expense reports submitted by these users this week from the task table, and displays the task lists corresponding to these expense reports on the UI; S43. Identify whether the user's input is a re-approval query request or an approval request; If it is an approval request, go to step S5; When the query request is to be reviewed again, the process goes to step S44; S44. The re-approval query request is integrated with the previous approval query request and used as the approval query request, and returns to step S32; For example, when there are too many expense reports submitted by users in the Finance Department this week, a second query is performed to "find the expense reports with high priority submitted by the Finance Department today". The system records that this query adds the task query variable "high priority", integrates it with the first task query variable, and compares and tracks it, and performs the approval query again; Or, if the user inputs: "pass the third task", it is identified as an approval request; It should be noted that the task information source is determined and queried according to the task query variable, and the tasks that meet the conditions are displayed to the user; by identifying whether the user input is a re-approval query request or an approval request, the re-query request is integrated and processed, thereby ensuring the coherent execution of the user query process and facilitating the user to further operate according to the query results; S5. Based on the task queue, respond to the user's approval request, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; the specific steps of step S5 are as follows: S51. Based on the task queue, identify the user's approval request and respond to it, and provide it to the natural language processing module of the large model; S52. The natural language processing module converts the user's approval request into natural language form and extracts the approval action and target task; Exemplarily, the natural language processing module identifies as follows: Approval Action: Pass Objectives: The third task S53. Generate confirmation prompts based on the approval action and target task, and return to the user; For example, the user is prompted through language or interface: “Do you want to pass the third task?” S54 responds to the user's confirmation of the confirmation prompt word and checks whether the approval conditions are met; If yes, go to step S57; If not, proceed to step S55; Exemplarily, after the user confirms, the system checks whether the approval conditions are met, such as whether it is necessary to fill in the approval opinion; S55. Generate approval prompt suggestions based on approval conditions and user behavior habits in the user behavior rule library; S56. respond to the user's feedback on the approval prompt suggestion and check whether the approval conditions are met; If yes, go to step S57; If not, return to step S55; S57. Execute the approval action on the target task in combination with the approval conditions to complete the approval operation; For example, if the approval conditions are met, the system directly performs the approval operation and updates the task status to "passed"; if the user is required to fill in the approval comments, the system prompts the user to enter the comments and performs the approval operation after the user completes the input; It should be noted that the entire process of natural language processing, extraction of approval actions and target tasks, generation and return of confirmation prompt words, inspection and satisfaction of approval conditions, and execution of approval actions ensures accurate and efficient execution of the approval process; by combining the user behavior rule library to generate approval prompt suggestions and making adjustments in response to user feedback, the intelligence and personalization of the approval system are improved; S6. Record the user's approval behavior process to the user behavior rule library; the specific steps of step S6 are as follows: S61. Analyze multiple approval query requests in the user's current approval behavior; S62. Identify the task query variables that have been increased or decreased in the subsequent approval query request compared to the previous approval query request, and mark the query using the condition type, and then add it to the user behavior habits of the user behavior rule library; Exemplarily, in the process of creating a mapping relationship between a query natural language and a task query variable, the user behavior habits are recorded, such as whether the user makes a second or third query after obtaining a query result based on the first query variable, and the variable conditions of each query are recorded, and compared and tracked with the last query variable respectively, and then, under the condition of confirming that it is a similar query, the difference between the last query variable and the previous query variables is compared, and the deleted query variables are marked as less used conditions for such queries, and the query variables added between the first and last time are marked as commonly used conditions. These query conditions can be associated with the user to ensure the user's usage habits to a greater extent; S63. Count the operations of the user on the approval conditions before executing the approval action in this approval behavior, and add the statistical results to the user behavior habits in the user behavior rule library; For example, when the approval condition requires filling in approval opinions, since the user's previous behavior habits are recorded in the user behavior rule library, if the user has a habit of simply clicking a button to approve before, the approval system will guide the user to fill in the approval opinions; if there is a record of filling in relevant opinions before, the approval system may give a reference approval result in advance for the user's reference based on the historical records and the user behavior rule library conditions; It should be noted that by analyzing the multiple approval query requests made by the user in this approval behavior, as well as counting the operations on the approval conditions before executing the approval action, and adding the results to the user behavior habits in the user behavior rule library, a data basis is provided for the approval system to learn user habits, thereby improving approval efficiency and user experience.

[0028] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0029] like Figure 2 As shown, the following is an embodiment of the process approval intelligent interaction system based on a big model provided by the embodiment of the present disclosure. This system and the process approval intelligent interaction method based on a big model in the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the process approval intelligent interaction system based on a big model, please refer to the embodiment of the process approval intelligent interaction method based on a big model.

[0030] The system includes: A user behavior rule base building module is used to create a user behavior rule base, store the mapping relationship between query natural language and task query variables, and record user behavior habits; An information source association relationship establishment module is used to determine a table related to task approval as a task information source and establish an association relationship between the task table and the task information source; The task query variable generation module is used to respond to the user's approval query request, convert the approval query request into query natural language through the big model, and generate task query variables according to the mapping relationship between the query natural language and the task query variable; The approval task query module is used to execute task query variables based on the association between the task table and the task information source, query all tasks that meet the conditions, generate a task queue, and display it to the user; The approval task execution module is used to respond to the user's approval request based on the task queue, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; Add a module to the user behavior rule library to record the user's approval behavior process in the user behavior rule library.

[0031] This embodiment realizes the full process automation from user behavior rule base establishment, information source association relationship management, task query variable generation, task query to approval operation execution and user behavior recording through the interaction of user behavior rule base establishment module, information source association relationship establishment module, task query variable generation module, approval task query module, approval task execution module and user behavior rule base addition module.

[0032] The intelligent interactive method for process approval based on a large model provided in the embodiment of the present application can be applied to electronic devices. Those skilled in the art will appreciate that the electronic device structure involved in the embodiment of the present invention does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown, or combine certain components, or arrange different components. In an embodiment of the present invention, electronic devices include but are not limited to laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the embodiments of the present application described and / or required herein.

[0033] The electronic device may include a processor, an external memory interface, an internal memory, a universal serial bus (USB) interface, a charging management module, a power management module, a battery, a wireless communication module, an audio module, a speaker, a microphone, a sensor module, buttons, a camera, a display, and a SIM card interface, etc.

[0034] It is to be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the electronic device. In other embodiments of the present application, the electronic device may include more or fewer components than shown in the figure, or combine certain components, or split certain components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0035] The processor may include one or more processing units, for example, the processor may include a central processing unit (CPU), an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a memory, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Among them, different processing units may be independent devices or integrated into one or more processors.

[0036] The processor can be the nerve center and command center of the electronic device. The controller can generate an operation control signal according to the instruction operation code and timing signal to complete the control of fetching and executing instructions.

[0037] A memory may also be provided in the processor for storing instructions and data. In some embodiments, the memory in the processor is a cache memory. The memory may store instructions or data that the processor has just used or is cyclically used. If the processor needs to use the instruction or data again, it may be directly called from the memory. This avoids repeated access, reduces the waiting time of the processor, and thus improves system efficiency.

[0038] The above-mentioned electronic device implements the intelligent interactive method of process approval based on a big model of the present application, which creates a user behavior rule library for the task center, stores the mapping relationship between natural language and task query variables, and records user behavior habits; determines the table related to task approval as the task information source, and establishes the association relationship between the task table and the task information source; responds to the user's approval query request, and converts it into a query natural language through a big model, and then generates the required task query variables based on the mapping relationship between the natural language and the task query variable; based on the required task query variables, and in combination with the association relationship between the task table and the relevant information source, queries all tasks that meet the conditions and displays them to the user; responds to the user's approval request, and converts it into approval natural language and parses it through a big model, determines the approval action and target task, checks the approval conditions and generates approval prompts and suggestions based on the user's behavior habits, and executes the approval action after the approval conditions are met; the technical solution of recording the user's approval behavior process into the user behavior rule library achieves the beneficial effects of improving approval efficiency, reducing user operation burden, optimizing user approval experience, and enhancing the intelligence of the approval process.

[0039] The storage medium provided in the present application stores a program product that can implement an intelligent interactive method for process approval based on a large model.

[0040] The intelligent interactive method for process approval based on the big model includes: creating a user behavior rule library for the task center, storing the mapping relationship between natural language and task query variables, and recording user behavior habits; determining the table related to task approval as the task information source, and establishing the association relationship between the task table and the task information source; responding to the user's approval query request, and converting it into query natural language through the big model, and then generating the required task query variables based on the mapping relationship between the natural language and the task query variable; querying all tasks that meet the conditions based on the required task query variables and the association relationship between the task table and the relevant information source and displaying them to the user; responding to the user's approval request, and converting it into approval natural language and parsing it through the big model, determining the approval action and target task, checking the approval conditions and generating approval prompt suggestions based on the user's behavior habits, and executing the approval action after the approval conditions are met; recording the user's approval behavior process in the user behavior rule library.

[0041] In some possible implementations, the big model-based process approval intelligent interactive method disclosed herein can be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps described in the above “Exemplary Method” section of this specification according to various exemplary implementations of the present disclosure.

[0042] The storage medium of the present disclosure can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, a system, device or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0043] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An intelligent interactive method for process approval based on a large model, characterized in that: The steps include: S1. Create a user behavior rule library to store the mapping relationship between query natural language and task query variables, and record user behavior habits; S2. Determine the table related to task approval as the task information source, and establish the association relationship between the task table and the task information source; S3. respond to the user's approval query request, and convert the approval query request into query natural language through the large model, and generate task query variables according to the mapping relationship between the query natural language and the task query variables; S4. executing the task query variable based on the association between the task table and the task information source, querying all tasks that meet the conditions, generating a task queue, and displaying it to the user; S5. Based on the task queue, respond to the user's approval request, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; S6. Record the user's approval behavior process into the user behavior rule library.

2. The intelligent interactive method for process approval based on a large model according to claim 1 is characterized in that: The specific steps of step S1 are as follows: S11. Define the task information fields on the task table; S12. Count the tasks in the task table according to the application scenario, and count the natural language queries in each scenario; S13. Extract keywords from the query natural language of each type of scenario in combination with the task information field defined in the task table to construct a task query variable; S14. Establish a mapping relationship between the query natural language and the task query variable, and save it in the user behavior rule library; S15. Analyze the user's historical query behavior to identify the user's behavior habits, and save them to the user behavior rule library.

3. The intelligent interactive method for process approval based on a large model according to claim 2 is characterized in that: The specific steps of step S2 are as follows: S21. Storing the task information corresponding to the task information field in the task table; S22. Determine the table related to task approval as the task information source; S23. Add the fields related to task approval in each task information source to the task table, and establish an association relationship between the task table and each task information source.

4. The intelligent interactive method for process approval based on a large model according to claim 3 is characterized in that: The specific steps of step S3 are as follows: S31. Identify the user's approval query request and respond to it, and provide it to the natural language processing module of the large model; S32. The natural language processing module converts the user's approval query request into natural language form, and obtains the task query variable and the corresponding variable value according to the mapping relationship between the natural language and the task query variable in the user behavior rule library.

5. The intelligent interactive method for process approval based on a large model according to claim 4 is characterized in that: The specific steps of step S4 are as follows: S41. Determine the required task information source according to the task query variable; S42. Query the task table according to the variable value of the task query variable and the required task information source, determine all tasks that meet the conditions, generate a task queue, and display it to the user; S43. Identify whether the user's input is a re-approval query request or an approval request; If it is an approval request, go to step S5; When the query request is to be reviewed again, the process goes to step S44; S44. Integrate the re-approval query request with the previous approval query request and use it as the approval query request, and return to step S32.

6. The intelligent interactive method for process approval based on a large model according to claim 5 is characterized in that: The specific steps of step S5 are as follows: S51. Based on the task queue, identify the user's approval request and respond to it, and provide it to the natural language processing module of the large model; S52. The natural language processing module converts the user's approval request into natural language form and extracts the approval action and target task; S53. Generate confirmation prompts based on the approval action and target task, and return to the user; S54 responds to the user's confirmation of the confirmation prompt word and checks whether the approval conditions are met; If yes, go to step S57; If not, proceed to step S55; S55. Generate approval prompt suggestions based on approval conditions and user behavior habits in the user behavior rule library; S56. respond to the user's feedback on the approval prompt suggestion and check whether the approval conditions are met; If yes, go to step S57; If not, return to step S55; S57. Execute the approval action on the target task in combination with the approval conditions to complete the approval operation.

7. The intelligent interactive method for process approval based on a large model according to claim 6 is characterized in that: The specific steps of step S6 are as follows: S61. Analyze multiple approval query requests in the user's current approval behavior; S62. Identify the task query variables that have been increased or decreased in the subsequent approval query request compared to the previous approval query request, and mark the query using the condition type, and then add it to the user behavior habits of the user behavior rule library; S63. Count the operations on the approval conditions before the user performs the approval action in this approval behavior, and add the statistical results to the user behavior habits in the user behavior rule library.

8. An intelligent interactive system for process approval based on a large model, characterized in that: include: A user behavior rule base building module is used to create a user behavior rule base, store the mapping relationship between query natural language and task query variables, and record user behavior habits; An information source association relationship establishment module is used to determine a table related to task approval as a task information source and establish an association relationship between the task table and the task information source; The task query variable generation module is used to respond to the user's approval query request, convert the approval query request into query natural language through the big model, and generate task query variables according to the mapping relationship between the query natural language and the task query variable; The approval task query module is used to execute task query variables based on the association between the task table and the task information source, query all tasks that meet the conditions, generate a task queue, and display it to the user; The approval task execution module is used to respond to the user's approval request based on the task queue, convert it into approval natural language through the big model and parse it, determine the approval action and target task, and execute the approval action; Add a module to the user behavior rule library to record the user's approval behavior process in the user behavior rule library.

9. An electronic device, characterized in that: The method 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 intelligent interactive method for process approval based on a large model as described in any one of claims 1 to 7 when executing the program.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the intelligent interactive method for process approval based on a large model as described in any one of claims 1 to 7 are implemented.

Images