Control method of approval process, computing device and readable storage medium

Abstract

The invention discloses an approval process control method, a computing device and a computer readable storage medium, and the method comprises the steps: inputting target task information into a pre-trained learning model, so as to output target approval path information of an approval process corresponding to the target task information, the learning model is obtained after model training according to a training set, and approval data in the training set comprises task information and approval path information; and according to the target approval path information, configuring an approval process corresponding to the target task information in an approval system. Thus, according to the technical scheme of the application, training of the learning model can be completed through the pre-constructed training set, so that the appropriate approval path can be automatically designed for the target task information through the learning model, the configuration difficulty of the approval path of the approval process can be reduced, the configuration efficiency of the approval process can be improved, and manpower can be saved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method for controlling an approval process, a computing device, and a computer-readable storage medium. Background Art

[0002] The current approval process mainly includes two forms, one is the online approval process and the other is the offline approval process. Among them, the offline approval process usually involves multiple departments and management levels of review, and the specific process includes: (1) Application submission: the applicant fills out a paper application form; (2) Approval process: the application form is submitted to the relevant department for approval at each level, and the approval personnel at each level need to pass the approved application form to the next level of approver until the approval process is completed. Among them, the online approval process is a form commonly used by enterprises or units with the deepening of digital transformation of enterprises or units. The online approval process has gradually replaced the offline approval process; the online approval process generally exists in the approval system through information technology; the specific process of the online approval process includes: (1) Electronic application submission: the applicant submits the application form online through the approval system, and the system automatically assigns the application form to the relevant approver; (2) Multi-level approval: the approval system automatically assigns a fixed approval path based on the preset rules, and the approvers at all levels can view and process the approval tasks online; (3) Approval task tracking and archiving: after the approval is completed, the system automatically archives all process records of the approval process. Compared with the traditional offline approval process, the current online approval process can ensure the legality and rationality of the approval, while the efficiency has been greatly improved.

[0003] However, the configuration of the approval path in the current online approval process relies on manual configuration, which makes the configuration of the approval path in the online approval process labor-intensive and complicated. Therefore, how to provide an efficient configuration method for the approval path in the approval process is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0004] The purpose of this application is to provide a control method, computing device and computer-readable storage medium for an approval process, which can provide an efficient configuration method for the approval path in the approval process to improve the construction and optimization efficiency of the approval process.

[0005] To achieve the above objectives: In the first aspect, an embodiment of the present application provides a method for controlling an approval process, including: inputting target task information into a pre-trained learning model to output target approval path information of the approval process corresponding to the target task information, the learning model being obtained after model training based on a training set, and the approval data in the training set including task information and approval path information; according to the target approval path information, configuring the approval process corresponding to the target task information in the approval system.

[0006] In one embodiment, before the step of inputting the target task information into a pre-trained learning model to output the target approval path information of the approval process corresponding to the target task information, it includes: using a training set to perform model training on the Q-learning reinforcement learning network to obtain a learning model.

[0007] In one embodiment, the control method of the approval process provided by the present application also includes: obtaining a historical approval data set based on historical approval data corresponding to at least one approval process under an approval category in the approval system; performing clustering processing based on characteristic parameters of node information in the historical approval data set to obtain at least one node clustering result, wherein the characteristic parameters include at least one of node identification information, node processing time, and node load; performing analysis and processing based on all node clustering results to determine the target approval node to be optimized, and executing an optimization strategy for the target approval node.

[0008] In one embodiment, analysis and processing are performed based on all node clustering results to determine the target approval node to be optimized, and the steps of executing the optimization strategy for the target approval node include: selecting a first clustering result that meets the labor efficiency optimization conditions from all node clustering results, and executing the labor efficiency optimization strategy for the target approval node corresponding to the first clustering result; and / or, selecting at least two second clustering results that meet the node merging conditions from all node clustering results, and executing merging control for the approval nodes corresponding to the at least two second clustering results.

[0009] In one embodiment, the human efficacy optimization condition includes at least one of the following: The node processing time corresponding to all or part of the node information in the first clustering result is greater than the preset time; The node loads corresponding to all or part of the node information in the first clustering result are all greater than the load threshold.

[0010] In one embodiment, the human efficiency optimization strategy includes at least one of the following: For the approval node corresponding to the first clustering result, the approval personnel configuration control corresponding to the senior functional personnel is performed to replace the original approval personnel in the approval node corresponding to the first clustering result with the senior functional personnel; For the approval node corresponding to the first clustering result, personnel addition control is performed to add new functional personnel based on the original approval personnel in the approval node corresponding to the first clustering result; The personnel information of the approving personnel is obtained according to the approval node corresponding to the first clustering result, and personnel training control is performed according to the personnel information.

[0011] In one embodiment, the personnel information of the approving personnel is obtained according to the approval node corresponding to the first clustering result, and personnel training control is performed according to the personnel information, including: obtaining the personnel information of the approving personnel according to the approval node corresponding to the first clustering result; determining the number of trainings according to the professional seniority in the personnel information; generating a training task according to the review operation guidance information and the number of trainings associated with the approval node corresponding to the first clustering result; and sending the training task to the user terminal corresponding to the notification receiving identifier according to the notification receiving identifier in the personnel information.

[0012] In one embodiment, the node information also includes node sequence information, and the characteristic parameters also include node approval results.

[0013] In one embodiment, the node merging condition includes at least one of the following: The node identification information corresponding to at least two second clustering results are different from each other; According to the node sequence information respectively corresponding to the at least two second clustering results, determining whether the approval nodes respectively corresponding to the at least two clustering results have a sequence on the approval path; The node approval results corresponding to at least two second clustering results are the same.

[0014] In a second aspect, an embodiment of the present application provides a computing device, comprising: a processor and a memory storing a computer program, and when the processor runs the computer program, the steps of the control method of the approval process described in any of the above items are implemented.

[0015] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the control method of the approval process described in any of the above items are performed.

[0016] The embodiment of the present application provides a control method, a computing device and a computer-readable storage medium for an approval process, wherein the control method for the approval process includes: inputting target task information into a pre-trained learning model to output target approval path information of the approval process corresponding to the target task information, the learning model is obtained after model training according to a training set, and the approval data in the training set includes task information and approval path information; according to the target approval path information, the approval process corresponding to the target task information is configured in the approval system. In this way, the technical solution of the present application can complete the training of the learning model through a pre-constructed training set, so that the target task information can be automatically designed through the learning model. The appropriate approval path can reduce the configuration difficulty of the approval path of the approval process, improve the configuration efficiency of the approval process, and save manpower. Therefore, the technical solution of the present application can provide an efficient configuration method for the approval path in the approval process to improve the construction and optimization efficiency of the approval process. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application. In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments are briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor.

[0018] Figure 1 It is a flowchart of the control method of the approval process provided in the embodiment of the present application.

[0019] Figure 2 This is a schematic diagram of the framework of the approval system of this application example.

[0020] Figure 3 It is a schematic diagram of the structure of the computing device provided in an embodiment of the present application.

[0021] The realization of the purpose, functional features and advantages of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings. The above-mentioned drawings have shown clear embodiments of this application, which will be described in more detail later. These drawings and textual descriptions are not intended to limit the scope of the concept of this application in any way, but to illustrate the concept of this application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0022] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0023] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element. In addition, components, features, and elements with the same name in different embodiments of the present application may have the same meaning or different meanings, and their specific meanings need to be determined by their explanation in the specific embodiment or further combined with the context of the specific embodiment.

[0024] It should be understood that, although the terms first, second, third, etc. may be used to describe various information in this article, these information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of this article, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "at the time of..." or "when..." or "in response to determination". Furthermore, as used in this article, the singular forms "one", "one" and "the" are intended to also include plural forms, unless there is an opposite indication in the context. It should be further understood that the terms "comprising", "including" indicate that there are described features, steps, operations, elements, components, projects, kinds, and / or groups, but do not exclude the existence, occurrence or addition of one or more other features, steps, operations, elements, components, projects, kinds, and / or groups. The terms "or" and "and / or" used herein are interpreted as inclusive, or mean any one or any combination. Thus, “A, B, or C” or “A, B and / or C” means “any of the following: A; B; C; A and B; A and C; B and C; A, B, and C.” An exception to this definition will occur only when a combination of elements, functions, steps, or operations are inherently mutually exclusive in some manner.

[0025] It should be understood that, although the various steps in the flowchart in the embodiment of the present application are displayed in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and it can be performed in other orders. Moreover, at least a portion of the steps in the figure may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and their execution order is not necessarily performed in sequence, but can be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0026] It should be noted that in this article, step codes such as S1 and S2 are used for the purpose of expressing the corresponding content more clearly and concisely, and do not constitute a substantial limitation on the order. When implementing the step, those skilled in the art may execute S2 first and then S1, etc., but these should all be within the scope of protection of this application.

[0027] It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0028] In the subsequent description, the suffixes such as "module", "component" or "unit" used to represent elements are only used to facilitate the description of the present application, and have no specific meanings. Therefore, "module", "component" or "unit" can be used in a mixed manner.

[0029] See also Figure 1 , which is a control method for an approval process provided in an embodiment of the present application. The method can be executed by a computing device provided in an embodiment of the present application, and the computing device can be implemented in software and / or hardware.

[0030] This embodiment provides a method for controlling an approval process, including the following steps (for example, step S1 to step S2): Step S1: Input the target task information into a pre-trained learning model to output the target approval path information of the approval process corresponding to the target task information. The learning model is obtained after model training based on the training set. The approval data in the training set includes task information and approval path information.

[0031] In one embodiment, the target task information may include at least one of the following information that can affect the generation of the approval path, such as approval category information, a sub-category under an approval category (or task type), and a task description.

[0032] In one embodiment, the target task information can be new task information for the approval system, or it can be original task information in the approval system. Optionally, the new task information corresponds to a new approval process. In this way, the technical solution of this embodiment can quickly generate the target approval path information of the corresponding approval process for the new task information through the learning model, so as to improve the efficiency of building a new approval process. Optionally, the original task information corresponds to an existing approval process in the approval system. In this way, the technical solution of this embodiment can quickly generate the target approval path information of the corresponding existing approval process for the original task information through the learning model, so as to replace the original approval path information of the existing approval process with the target approval path information, thereby realizing the approval path optimization function of the existing approval process, and can improve the optimization efficiency of the approval path of the existing approval process.

[0033] In one implementation, the approval category information may be used to determine the approval category to which the currently configured approval process belongs.

[0034] In one embodiment, a variety of different approval processes can be created under an approval category, and each approval process can correspond to a subdivision type under the approval category. In other words, different subdivision types under the approval category can correspond to the same or different approval processes. Optionally, different approval processes can be manifested as different approval paths configured in the approval process, and / or different approval nodes configured in the approval process. Taking the purchase approval category as an example, the approval path of the purchase approval process corresponding to the production material category can include approval node A1, approval node B, approval node C, and approval node D in sequence, and approval node B is configured with approval personnel b1 and approval personnel b2; wherein, the approval path of the purchase approval process corresponding to the office supplies category can include approval node A2, approval node B, and approval node D in sequence, and approval node B is only configured with approval personnel B2.

[0035] In one embodiment, a sub-category under an approval category may represent a sub-category under an approval category. For example, a procurement sub-category under a procurement category may include production material procurement, office supplies procurement, etc.; another example, a requisition sub-category under a logistics item requisition category may include office supplies requisition, cleaning supplies requisition.

[0036] In one implementation, the target approval path information may represent each approval node in an approval process, as well as the flow logic between each approval node and related information.

[0037] In one embodiment, the training set may represent a set of approval data. In other words, the training set may include one or more approval data.

[0038] In one embodiment, the approval data may be used to analyze, optimize and understand the configuration, performance and other characteristics of the approval process (e.g., the approval node of the approval process or the approval path of the approval process). Optionally, each piece of approval data may represent a complete approval event of an approval process. Optionally, each piece of approval data may record all relevant information from the beginning to the end of a single approval process and / or the process configuration information of a single approval process, including but not limited to task information and approval path information, and may also include node information.

[0039] In one embodiment, the approval data may include virtual approval data and / or real historical approval data.

[0040] In one embodiment, the virtual approval data can be approval data obtained by artificially simulating the configuration and approval process of the real approval process, which can make the parameters of the learning model converge quickly during model training to quickly complete model training. Optionally, the virtual approval data can be used as the training basis of the learning model in the early stage of the establishment of the approval system, so that the approval system with the function of automatically generating the approval path for the approval process can be quickly established and put into use.

[0041] In one embodiment, the real historical approval data may include offline historical approval data corresponding to the approval process used by a unit or organization, and online historical approval data corresponding to the approval process used by a unit or organization or circulated in the approval system. The technical solution of this embodiment can use real historical approval data to build a training set to train the learning model, so that the target approval path information finally obtained by the learning model according to the target task information output is more in line with the configuration habits or configuration regulations of the approval path of the approval process of a specific unit or organization, so it can ensure that the approval system has localized and personalized adaptation functions when creating an approval process.

[0042] In one embodiment, the information framework of the task information included in the approval data in the training set is the same as or similar to the information framework of the target task information. For example, the information framework of the target task information may include approval category information, subcategories under an approval category (or task types), task descriptions, etc., and the information framework of the task information included in the approval data in the training set also includes approval category information, subcategories under an approval category (or task types), task descriptions, etc., and only the specific content or parameters in the information framework are different, then it is considered that the information framework of the task information included in the approval data in the training set is the same as the information framework of the current task information.

[0043] In one embodiment, in order to facilitate understanding of the information framework of task information in real historical approval data, the following examples are given for reference: the task information of a procurement approval process under a procurement category may include task ID, procurement sub-category (such as production materials, office supplies, etc.), budget amount, supplier information, task initiator ID, initiator department, task submission time, task description, etc.; for example, the task information of an item receipt approval process under a logistics item receipt category may include task ID, receipt sub-category (such as office supplies, cleaning supplies, etc.), cost amount, item list, task initiator ID, initiator department, task submission time, task description, task completion requirement time, etc.

[0044] In one embodiment, the approval path information included in the approval data in the training set can also represent the various approval nodes in an approval process, as well as the flow logic between the various approval nodes and related information, compared with the target approval path information. Optionally, the approval path information included in the approval data in the training set can also represent the approval path evaluation of the approval process to which it corresponds (such as path processing time, path complexity, etc.), so as to facilitate the subsequent adjustment of the weights in the learning model.

[0045] In one embodiment, when the approval data in the training set is real historical approval data, the approval path information included therein may include path identification information (such as path ID), number of path nodes, node flow order, path processing time (or task processing time), task approval results and at least one of other features related to the approval path.

[0046] In one implementation, the path identification information may represent information that distinguishes one approval path from other approval paths.

[0047] In one implementation, the number of path nodes represents the number of approval nodes configured in an approval path.

[0048] In one implementation, the node flow sequence represents the flow logic between each approval node.

[0049] In one implementation, the path processing duration may represent the actual duration of the approval process of configuring an approval path from the time point when the approval starts to the time point when the approval ends.

[0050] In one implementation, the task approval result may represent an approval completion form of an approval process, including but not limited to one of the approval completion forms such as completion and filing, voiding and filing, and rejection.

[0051] In one embodiment, the node information included in the approval data in the training set may include at least one of node identification information (e.g., node ID), node sequence information, node role information, node processing time, node processing duration, node approval results, node load, and other node-related features.

[0052] In one implementation, the node identification information may represent information for distinguishing an approval node from other approval nodes.

[0053] In one embodiment, the node sequence information may represent the sequence position of the approval node among all the approval nodes in the approval process to which it belongs, or represent the sequence position of the approval node among all the approval nodes in the approval category. Optionally, the sequence position relationship between the approval node and other approval nodes in an approval process or an approval category may be determined based on the node sequence information.

[0054] In one embodiment, the node role information may represent the description or identification of the duties, responsibilities and / or functions of an individual or group associated with an approval process or an approval category. For example, if the approval node corresponds to compliance approval, the node role information corresponds to the responsibilities and / or functional description of the legal department; for another example, if the approval node corresponds to financial approval, the node role information corresponds to the responsibilities and / or functional description of the financial department.

[0055] In one implementation, the task processing time may represent the specific time of submitting a node approval result of an approval flow under the current approval node.

[0056] In one implementation, the node processing time may represent the actual time elapsed from the time point when an approval node receives an approval task of an approval process to the time point when the node approval result is submitted.

[0057] In one embodiment, the node approval result may include, but is not limited to, one of the results such as approval, rejection (e.g., return to the initial approval node of the approval process), return (e.g., return to the previous approval node of the current approval node or return to any set previous approval node).

[0058] In one implementation, the node load may represent the number of other approval processes to be approved in an approval node when the approval node is approving the current approval process.

[0059] In one embodiment, the construction method of the training set may include: adding online historical approval data corresponding to at least one approval process under an approval category to the historical approval data set, and / or converting offline historical approval data corresponding to at least one approval process under an approval category into historical approval data in a preset form, and then adding the preset form of historical approval data to the historical approval data set; selecting historical approval data from the historical approval data set and adding it to the training set, wherein the selected historical approval data may be historical approval data of an approval process that meets the high-quality condition. In this way, the technical solution of this embodiment can use the training set constructed with high-quality historical approval data of the approval process to train the learning model, so that the final learning model can output high-quality target approval path information according to the target task information.

[0060] In one embodiment, the historical approval data may also include a case closing score, which may be obtained after the approval process is completed and the applicant or other relevant personnel of the approval process performs a scoring operation.

[0061] In one embodiment, the high-quality conditions include but are not limited to at least one of the following: The task processing time in the historical approval data of the approval process is less than the preset task time; The number of path nodes in the historical approval data of the approval process is less than the preset number of nodes; The case closing scores in the historical approval data of the approval process meet the preset scores that represent better satisfaction.

[0062] In one embodiment, after converting offline historical approval data corresponding to at least one approval process under an approval category into historical approval data in a preset form, the step of adding the historical approval data in the preset form to the historical approval data set may include: photographing the paper application form carrying the offline historical approval data to obtain the application form photo; recognizing the text information in the application form photo by image recognition technology to obtain task information and personnel signature information; determining at least one approval process template under an approval category according to the task information; determining the personnel information of the approval personnel according to the personnel signature information, and determining the configuration information of at least one approval node according to the personnel function or the department function corresponding to the personnel information; filling the information in the approval process template according to the configuration information and task information of at least one approval node to obtain the historical approval data in a preset form; adding the historical approval data in the preset form to the historical approval data set. In this way, the technical solution of this embodiment can convert offline historical approval data into historical approval data in a preset form that can be added to the historical approval data set through image recognition technology and approval process template matching technology. Therefore, the technical solution of this embodiment can improve the efficiency of collecting and entering offline historical approval data, and for the scenario of data import in the early stage of building an approval system, it can greatly improve the efficiency of building and launching the approval system.

[0063] The technical solution of this embodiment can collect real historical approval data of one or more approval processes under an approval category based on the approval category, so that the real historical approval data under an approval category can be processed later to construct a training set for an approval category, and then a learning model corresponding to the approval category can be constructed based on the training set, so as to realize the automated design of the approval path of one or more approval processes under an approval category and improve the efficiency of establishing the approval process.

[0064] In one embodiment, before the step of inputting the target task information into a pre-trained learning model to output the target approval path information of the approval process corresponding to the target task information, it includes: using the training set to train the Q-learning reinforcement learning network to obtain the learning model. In this way, the technical solution of this embodiment can use the Q-learning reinforcement learning network to simulate the decision-making process of the approval process, automatically learn the optimal approval node configuration and sequence, so that the optimal approval path information can be output according to the imported target task information. Therefore, the technical solution of this embodiment uses Q-learning reinforcement learning to realize the construction function or optimization function of the approval path of the approval process corresponding to the target task information.

[0065] In one embodiment, the approval data in the training set is used to train a Q-learning reinforcement learning network to learn the configuration of the approval path, thereby obtaining a learning model, which can be performed according to the following steps: (1) Data preprocessing: First, the approval data is cleaned and formatted to ensure data consistency and availability. This may include removing duplicate records, processing missing values, converting data formats, etc.

[0066] (2) Feature extraction: Useful features are extracted from the approval data. These features may include task features, node features, and / or process features. These features will be input into the reinforcement learning model as part of the state space.

[0067] (3) Define states and actions: State: In Q-learning, state is the basis for algorithm evaluation and decision making. In the approval process, state can be defined as a specific stage of the approval process, including the current node, the current approver, and the task information to be approved.

[0068] Action: An action is an operation that an agent can perform. In an approval process, actions may include approval, rejection, request for more information, forwarding to the next approval node, etc.

[0069] (4) Initialize Q-table: Create a Q-table to store the value (Q-value) of each state-action pair. This table can be initially initialized to zero or a small random number.

[0070] (5) Define the reward function: Design a reward function to evaluate the quality of actions. The reward function should be able to reflect the efficiency and effectiveness of the approval process. For example, quick approval can be given positive rewards, while unnecessary delays can be given negative rewards.

[0071] (6) Training model: Use the approval data as the training set and train the model by simulating each step of the approval process.

[0072] In each state, actions are selected according to the Q-table and ε-greedy strategies. Among them, the ε-greedy strategy is a commonly used exploration and exploitation strategy in reinforcement learning. In this strategy, the agent will choose random actions with a certain probability ε (usually a very small positive number) to explore the environment when making decisions; or choose the current optimal action with a probability of 1-ε to exploit known information.

[0073] After executing the action, observe the new state and the reward received, and update Formula; where The formula includes operators that represent the learning rate, discount factor, reward after performing the action, new state, and possible actions in the new state.

[0074] (7) Model evaluation and optimization: Evaluate the performance of the model through the test set, and adjust the model parameters such as learning rate, discount factor, exploration rate, etc. based on the evaluation results.

[0075] (8) Model deployment: Deploy the trained learning model to the approval system so that it can automatically output the recommended target approval path information based on the target task information.

[0076] (9) Monitoring and maintenance: After the learning model is deployed, its performance is continuously monitored, and necessary maintenance and updates are performed based on feedback from actual applications.

[0077] Through the above process, the Q-learning algorithm can learn what actions should be taken under different approval states to optimize the efficiency and effectiveness of the entire approval process.

[0078] In one embodiment, in addition to the Q-learning reinforcement learning network, the reinforcement learning network for constructing the learning model may also use more mature learning networks such as Deep Q-Network (DQN) and Policy Gradients.

[0079] In one embodiment, feature extraction is performed on the approval data in the training set to extract task features, node features and / or process features, etc. The extracted features are input into the reinforcement learning network to adjust the network parameters in the reinforcement learning network to obtain a learning model.

[0080] In one embodiment, the task characteristics correspond to the task information, including a sub-category (or task type) under an approval category and other information that affects the configuration of the approval process. For example, to configure a purchase approval process violation, the task characteristics may include the purchase task type, budget amount, etc.

[0081] In one embodiment, the node feature may represent the approval work evaluation information of each approval node when the approval process corresponding to the approval data is completed. Optionally, the node feature information may include extracting at least one of the processing time of each approval node, the node pass rate of each approval node, and the decision accuracy of each approval node.

[0082] In one embodiment, the process feature may represent the approval work evaluation information of the approval process corresponding to the approval data when the approval process is completed. Optionally, the process feature may be a feature obtained by analyzing the entire approval process, which may include the total number of nodes, total processing time, path complexity, task completion rate, etc.

[0083] Step S2: According to the target approval path information, the approval process corresponding to the target task information is configured in the approval system.

[0084] In one embodiment, step S2: configuring the approval process corresponding to the target task information in the approval system according to the target approval path information includes one of the following: When the target task information is new task information, a new approval process is created in the approval system and the approval path is configured according to the target approval path information; When the target task information is existing task information, an existing approval process matching the target task information is determined in the approval system, and the original approval path information of the existing approval process is replaced with the target approval path information to update the approval path of the existing approval process.

[0085] In this way, the technical solution of this embodiment can quickly generate the target approval path information of the corresponding approval process for the newly added task information through the learning model, so as to improve the efficiency of building a new approval process. In addition, it can also quickly generate the corresponding target approval path information of the existing approval process for the original task information through the learning model, so as to replace the original approval path information of the existing approval process with the target approval path information, thereby realizing the approval path optimization function of the existing approval process and improving the optimization efficiency of the approval path of the existing approval process.

[0086] In one embodiment, when the target approval path information includes node information, the approval system can complete the approval path configuration of the approval process for the target task information and the configuration of each approval node (for example, configuring the form, field, approval rules and approval personnel of the approval node, etc.) according to the target approval path information.

[0087] In one embodiment, when the target approval path information does not include node information, the approval system can complete the approval path configuration of the approval process for the target task information according to the target approval path information, and subsequently perform other node configuration operations to configure each approval node of the approval process.

[0088] In one embodiment, the control method of the approval process provided by this embodiment (for example, in step S2: after configuring the approval process corresponding to the target task information in the approval system according to the target approval path information) may also include: obtaining a historical approval data set based on historical approval data corresponding to at least one approval process under an approval category in the approval system; performing clustering processing based on characteristic parameters of the node information in the historical approval data set to obtain at least one node clustering result, wherein the characteristic parameters include at least one of node identification information, node processing time, and node load; performing analysis and processing based on all node clustering results to determine the target approval node to be optimized, and executing an optimization strategy for the target approval node. In this way, the technical solution of this embodiment can cluster the node information in the historical approval data of the approval process under an approval category configured in the approval system, so as to assign the node information with the same or similar characteristic parameters (such as node identification information, node processing time and / or node load) to the same cluster to obtain at least one clustering result, so as to quickly determine the target approval node that needs to be optimized for efficiency in the approval process under an approval category and execute the optimization strategy by analyzing at least one clustering result, so as to improve the approval efficiency of the approval process under an approval category in the approval system from the approval node level. Therefore, the technical solution of this embodiment can not only provide an efficient configuration method for the approval path in the approval process, so as to improve the approval system's construction and optimization efficiency of the approval process, but also provide an approval efficiency optimization method for the approval nodes in the approval process constructed and used in the approval system, so as to improve the overall approval efficiency of the approval process constructed and used in the approval system.

[0089] In one embodiment, the characteristic parameters of the node information may represent information selected from the node information and used as a clustering condition factor. It should be understood that the characteristic parameters may include, but are not limited to, one or more of identification information, node processing time, and node load, and may also include, for example, node approval results, node sequence information, etc.

[0090] In one embodiment, clustering is performed according to characteristic parameters of node information in the historical approval data set to obtain at least one node clustering result, which may include: clustering all node information in the historical approval data set by a clustering algorithm to assign node information whose characteristic parameters meet the clustering conditions to the same cluster to obtain at least one node clustering result. In this way, the technical solution of this embodiment can perform clustering on the node information in the real historical approval data by a clustering algorithm, so as to conveniently and quickly assign node information with the same or similar characteristic parameters (such as node identification information, node processing time and / or node load) to the same cluster to obtain at least one clustering result, so as to facilitate the subsequent analysis of at least one clustering result.

[0091] In one embodiment, the clustering condition may represent the similarity of characteristic parameters constraining different node information to perform various conditions for clustering.

[0092] In one embodiment, the clustering algorithm includes but is not limited to one of the K-means clustering algorithm, the K-medoids clustering algorithm, the k-sums clustering algorithm, etc. Among them, the K-means clustering algorithm is a simple and widely used clustering algorithm, which aims to divide data points into a predetermined number of clusters. The goal of the algorithm is to minimize the variance within the cluster, that is, the sum of the distances of each point to its cluster center. The following are the basic steps of the K-means algorithm: (1) Select the K value: determine the number of clusters to be formed; (2) Initialize the center: randomly select K data points (such as node information) as the initial cluster center (centroid); (3) Assign data points: assign each data point to the nearest cluster center to form K clusters; (4) Update the center: calculate the new center of each cluster, usually the mean of all points in the cluster; (5) Iterate optimization: repeat steps (3) and (4) until the cluster center no longer changes significantly or reaches a predetermined number of iterations; (4) Evaluate convergence, usually by comparing the changes in the positions of cluster centers in consecutive iterations; (5) Result evaluation: use indicators such as the silhouette coefficient and the Davies-Bouldin index to evaluate the quality of the clustering results. The advantages of the K-means algorithm include: the algorithm is simple and easy to implement; the clustering speed is fast, especially when the data set is not particularly large; in practical applications, the K-means clustering algorithm has a good clustering effect when applied to document clustering. Among them, K-medoids clustering algorithm: K-medoids algorithm is similar to K-means, but the center point of the cluster is selected as the actual data point, not the mean point. This makes K-medoids more robust to noise and outliers. Among them, k-sums clustering algorithm: This is an efficient clustering algorithm that can replace k-means, especially suitable for scenarios that are sensitive to initialization; k-sums algorithm divides the data set by optimizing the evaluation function, does not need to specify the number of clusters in advance, and can handle data sets of different densities and cluster sizes.

[0093] In one embodiment, analysis and processing are performed based on all node clustering results to determine the target approval node to be optimized, and the steps of executing the optimization strategy for the target approval node include: selecting a first clustering result that meets the labor efficiency optimization conditions from all node clustering results, and executing the labor efficiency optimization strategy for the target approval node corresponding to the first clustering result; and / or, selecting at least two second clustering results that meet the node merging conditions from all node clustering results, and executing merging control for the approval nodes corresponding to the at least two second clustering results.

[0094] In one embodiment, the human efficacy optimization condition includes at least one of the following: The node processing time corresponding to all or part of the node information in the first clustering result is greater than the preset time; The node loads corresponding to all or part of the node information in the first clustering result are all greater than the load threshold.

[0095] In one embodiment, the preset duration may be a preset maximum value of the node processing duration corresponding to the node information in the first clustering result. It should be understood that different node information may be set with the same or different preset durations.

[0096] In one implementation, the node processing durations corresponding to all or part of the node information in the first clustering result are all greater than a preset duration, which may indicate that the approval nodes corresponding to the node information in the first clustering result are approval nodes with low processing efficiency.

[0097] In one implementation, the load threshold may be a maximum number of other approval processes to be processed corresponding to the preset node information.

[0098] In one embodiment, the node loads corresponding to all or part of the node information in the first clustering result are greater than the load threshold, which can indicate that the approval node corresponding to the node information in the first clustering result is the approval node with the largest number of configurations (or key nodes).

[0099] In one embodiment, the node processing time corresponding to all or part of the node information in the first clustering result is greater than the preset time, and the node load corresponding to all or part of the node information in the first clustering result is greater than the load threshold, which can directly characterize the key nodes.

[0100] In one implementation, the human efficiency optimization strategy may represent a personnel configuration adjustment method for the approval node corresponding to the first clustering result, so as to improve the node approval efficiency.

[0101] In one embodiment, the human efficiency optimization strategy includes at least one of the following: For the approval node corresponding to the first clustering result, the approval personnel configuration control corresponding to the senior functional personnel is performed to replace the original approval personnel in the approval node corresponding to the first clustering result with the senior functional personnel; For the approval node corresponding to the first clustering result, personnel addition control is performed to add new functional personnel based on the original approval personnel in the approval node corresponding to the first clustering result; The personnel information of the approving personnel is obtained according to the approval node corresponding to the first clustering result, and personnel training control is performed according to the personnel information.

[0102] In one embodiment, a senior functional personnel can be characterized by the functions and / or responsibilities in the personnel information of the personnel matching the node role information of the approval node corresponding to the first clustering result, and the professional seniority or years of service in the personnel information exceeds the first preset years, and / or the professional title in the personnel information corresponds to a senior professional title.

[0103] In one embodiment, the approval node corresponding to the first clustering result is controlled by the configuration of the approval personnel corresponding to the senior functional personnel, so as to replace the original approval personnel in the approval node corresponding to the first clustering result with the senior functional personnel, which may include: when it is determined that the approval node corresponding to the first clustering result is an approval node with low processing efficiency or a key node, all the personnel information configured in the approval node corresponding to the first clustering result is obtained; when it is determined based on all the personnel information that the approval node corresponding to the first clustering result is configured with a junior functional personnel, the senior functional personnel is determined based on the node role information of the approval node corresponding to the first clustering result or the department information of the junior functional personnel; and the junior functional personnel in the approval node corresponding to the first clustering result is replaced with a senior functional personnel. In this way, the technical solution of this embodiment can directly retrieve the department information of the junior functional personnel to find the appropriate senior functional personnel, and complete the replacement of the approval personnel in the approval process.

[0104] In one embodiment, a junior functional personnel may represent that the functions and / or responsibilities in the personnel information of the personnel match the node role information of the approval node corresponding to the first clustering result, and the professional experience or years of employment in the personnel information is less than the second preset years, and / or the title in the personnel information corresponds to a junior title. The second preset years may be the same as or different from the first preset years.

[0105] The technical solution of this embodiment can realize the deployment of senior functional personnel to approval nodes or key nodes with low processing efficiency, thereby improving the approval efficiency of the node, and further achieving the purpose of improving the overall approval efficiency of the approval process.

[0106] In one implementation, the functional personnel may represent that the functions and / or responsibilities in the personnel information of the personnel match the node role information of the approval node corresponding to the first clustering result.

[0107] In one embodiment, the approval node corresponding to the first clustering result is controlled to increase personnel, so as to add new functional personnel on the basis of the original approval personnel in the approval node corresponding to the first clustering result. The step may include: when it is determined that the approval node corresponding to the first clustering result is an approval node with low processing efficiency or a key node, based on the node role information of the approval node corresponding to the first clustering result, obtaining the target functional personnel matching the node information, wherein the target functional personnel is different from the original approval personnel in the approval node corresponding to the first clustering result; configuring the target functional personnel on the basis of the original approval personnel in the approval node corresponding to the first clustering result to increase the number of approval personnel in the approval node corresponding to the first clustering result.

[0108] In one implementation, the personnel information of the target functional personnel and the personnel information of the original approval personnel may correspond to the same department information.

[0109] The technical solution of this embodiment can realize increasing the number of approval personnel for approval nodes or key nodes with low processing efficiency, thereby improving the approval efficiency of the node and controlling the node load, and further realizing the purpose of improving the overall approval efficiency of the approval process. Therefore, the technical solution of this embodiment can balance the task load of the approval personnel by optimizing the allocation of personnel at the approval nodes to avoid excessive tasks for the approval personnel at the approval nodes, thereby realizing the purpose of improving the overall approval efficiency of the approval process.

[0110] In one embodiment, the personnel information may include at least one of various information that can identify or characterize personnel capabilities, such as professional experience / years of service, functions, responsibilities, and notification reception identification.

[0111] In one embodiment, the notification receiving identifier may represent the communication identification information corresponding to the way in which a person receives various information. For example, if a person uses his or her mobile phone as a user terminal to receive various information, the notification receiving identifier may be a mobile phone number; for another example, if a person logs in to a user terminal of a network platform to receive various information, the notification receiving identifier may be the account number of the person logging in to the user terminal.

[0112] In one embodiment, the personnel information of the approving personnel is obtained according to the approval node corresponding to the first clustering result, and personnel training control is performed according to the personnel information, including: obtaining the personnel information of the approving personnel according to the approval node corresponding to the first clustering result; determining the number of trainings according to the professional seniority in the personnel information; generating a training task according to the review operation guidance information and the number of trainings associated with the approval node corresponding to the first clustering result; and sending the training task to the user terminal corresponding to the notification receiving identifier according to the notification receiving identifier in the personnel information.

[0113] In one embodiment, the audit operation guidance information may be at least one of audit operation guidance video information, audit operation guidance text information, audit operation guidance picture information, etc., which is used to convey correct and efficient audit operation points and skills to personnel.

[0114] The technical solution of this embodiment can arrange training tasks for the reviewers of the review nodes or key nodes with low processing efficiency, so that the reviewers can improve the review speed and review accuracy after learning the training tasks. Furthermore, the technical solution of this embodiment can determine the same or different training times according to the different professional years of the personnel, and can implement multiple trainings for junior functional personnel so that they can quickly master the key points and skills of the review operation to improve the review efficiency and review accuracy of the nodes, and can implement standardized reminders for the review behavior of senior functional personnel.

[0115] In one implementation, the node information may further include node sequence information, and the characteristic parameters may further include node approval results.

[0116] In one embodiment, the node merging condition may represent a condition for identifying nodes with duplicate approval logic and functions.

[0117] In one embodiment, the node merging condition includes, but is not limited to, at least one of the following: The node identification information corresponding to at least two second clustering results are different from each other; According to the node sequence information respectively corresponding to the at least two second clustering results, determining whether the approval nodes respectively corresponding to the at least two clustering results have a sequence on the approval path; The node approval results corresponding to at least two second clustering results are the same.

[0118] In one embodiment, the node identification information corresponding to at least two second clustering results are different from each other, and the node approval results corresponding to at least two second clustering results are the same, which can indicate that the approval conclusions of at least two different approval nodes highly overlap. Then, at least two different approval nodes are determined to be nodes with repeated approval logic and functions, so that merge control can be performed on the nodes with repeated approval logic and functions in the future to reduce approval nodes and shorten the approval path of the approval process, thereby improving the overall approval efficiency of the approval process.

[0119] In one embodiment, the node identification information corresponding to at least two second clustering results are different from each other, the node approval results corresponding to at least two second clustering results are the same, and the approval nodes corresponding to at least two clustering results are determined to have a sequence on the approval path based on the node sequence information corresponding to at least two second clustering results. This can indicate that the approval conclusions of at least two different approval nodes configured on a type of approval path are highly overlapped, and can more accurately determine that at least two different approval nodes configured on a type of approval path are nodes with repeated approval logic and functions, so that the nodes with repeated approval logic and functions can be subsequently merged and controlled to reduce approval nodes and shorten the approval path of the approval process, thereby improving the overall approval efficiency of the approval process.

[0120] The present embodiment provides a method for controlling an approval process, comprising the following steps (e.g., step S1 to step S2): step S1: inputting target task information into a pre-trained learning model to output target approval path information of the approval process corresponding to the target task information, the learning model is obtained after model training according to a training set, and the approval data in the training set includes task information and approval path information; step S2: configuring the approval process corresponding to the target task information in the approval system according to the target approval path information. In this way, the technical solution of the present embodiment can complete the training of the learning model through a pre-constructed training set, so that the learning model can be used to automatically design a suitable approval path for the target task information, thereby reducing the configuration difficulty of the approval path of the approval process, improving the configuration efficiency of the approval process, and saving manpower. Therefore, the technical solution of the present embodiment can provide an efficient configuration method for the approval path in the approval process to improve the construction and optimization efficiency of the approval process.

[0121] Based on the same inventive concept as the above embodiments, see Figure 2 , an approval system that applies the control method of the above-mentioned approval process is exemplified below to illustrate the above-mentioned embodiment.

[0122] The approval system provided in this example includes a data set construction module M1, an approval node optimization module M4, an optimization path integration module M3 and an approval path optimization module M4.

[0123] Among them, the functions of the dataset construction module M1 include: (1) Data collection: Collect application form data, approval node data, and approval path data of the approval process; (2) Data collation: collate task information, node information and / or approval path information based on application form data, approval node data and / or approval path data; (3) Data format conversion: Obtain historical approval data based on task information, node information and / or approval path information, and store it in a structured format (CVS) to obtain a historical approval data set.

[0124] Among them, the functions of the approval path optimization module M2 include: (1) Feature extraction: extracting input features from the training set based on the historical approval dataset; (2) Model construction: Import input features into the Q-learning reinforcement learning network for model training, so that the Q-learning reinforcement learning network learns the optimal approval path to obtain a learning model; (3) Path optimization: Import the target task information into the learning model to output the optimal target approval path information corresponding to the target task information.

[0125] Among them, the functions of the optimized path integration module M3 include: (1) Path integration: Embed the target task information and its corresponding target approval path information into the approval rules to generate the approval process and put it online.

[0126] Among them, the functions of the approval node optimization module M4 include: (1) Node data analysis: Use the K-means clustering algorithm to perform cluster analysis on the node information in the historical approval data set to identify key nodes, approval nodes with low processing efficiency, and / or approval nodes with repeated approval logic; (2) Node optimization: Implement human efficiency optimization strategies for key nodes and approval nodes with low processing efficiency, and implement merge control for approval nodes with repeated approval logic.

[0127] In summary, the technical solution provided by the above example can enable the approval system to have the function of optimizing the approval efficiency of the approval nodes in the approval process and the function of generating and optimizing the approval path for the approval task. The technical solution provided by the above example combines K-means clustering and Q-learning reinforcement learning, so that the approval system has the function of generating and optimizing the approval nodes and approval paths efficiently and intelligently.

[0128] Based on the same inventive concept as the above embodiments, the present application embodiment provides a computing device, such as Figure 3 As shown, the device includes: a processor 310 and a memory 311 storing a computer program; wherein, Figure 3 The processor 310 shown in the figure is not used to indicate that the number of the processor 310 is one, but is only used to indicate the positional relationship of the processor 310 relative to other devices. In actual applications, the number of the processor 310 may be one or more; similarly, Figure 3The memory 311 shown in the figure has the same meaning, that is, it is only used to refer to the position relationship of the memory 311 relative to other devices. In practical applications, the number of memories 311 can be one or more. When the processor 310 runs the computer program, the control method for the approval process applied to the above device is implemented.

[0129] The device may also include: at least one network interface 312. The various components in the device are coupled together via a bus system 313. It is understood that the bus system 313 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 313 also includes a power bus, a control bus, and a status signal bus. However, for the sake of clarity, Figure 3 Various buses are labeled as bus system 313.

[0130] The memory 311 may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. The non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a magnetic random access memory (FRAM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); the magnetic surface memory may be a disk memory or a tape memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), synchronous static random access memory (SSRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM, SyncLink Dynamic Random Access Memory), and direct RAMbus random access memory (DRRAM, Direct Rambus Random Access Memory).The memory 311 described in the embodiments of the present application is intended to include but is not limited to these and any other suitable types of memories.

[0131] The memory 311 in the embodiment of the present application is used to store various types of data to support the operation of the device. Examples of these data include: any computer program used to operate on the device, such as an operating system and an application program. Among them, the operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and handle hardware-based tasks. The application program may include various applications, such as a media player (MediaPlayer), a browser (Browser), etc., which are used to implement various application services. Here, the program that implements the method of the embodiment of the present application may be included in the application program.

[0132] Based on the same inventive concept as the above-mentioned embodiment, this embodiment further provides a computer-readable storage medium, in which a computer program is stored. The computer-readable storage medium may be a ferromagnetic random access memory (FRAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), a flash memory, a magnetic surface memory, an optical disc, or a compact disc read-only memory (CD-ROM); it may also be various devices including one or any combination of the above-mentioned memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc. When the computer program stored in the computer-readable storage medium is executed by the processor, the control method of the above-mentioned approval process is implemented. For the specific steps implemented when the computer program is executed by the processor, please refer to Figure 1 The control method of the approval process described in the illustrated embodiment will not be repeated here.

[0133] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0134] In this document, the terms "comprises," "comprising," or any other variations thereof, are intended to cover a non-exclusive inclusion of elements other than those listed and may also include additional elements not expressly listed.

[0135] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for controlling an approval process, characterized in that: include: Inputting target task information into a pre-trained learning model to output target approval path information of the approval process corresponding to the target task information, the learning model is obtained after model training based on a training set, and the approval data in the training set includes task information and approval path information; The approval process corresponding to the target task information is configured in the approval system according to the target approval path information.

2. The method according to claim 1, characterized in that Before the step of inputting the target task information into the pre-trained learning model to output the target approval path information of the approval process corresponding to the target task information, the method includes: The training set is used to perform model training on the Q-learning reinforcement learning network to obtain the learning model.

3. The method according to claim 1, characterized in that Also includes: Acquire a historical approval data set according to historical approval data corresponding to at least one approval process under an approval category in the approval system; Performing clustering processing according to characteristic parameters of the node information in the historical approval data set to obtain at least one node clustering result, wherein the characteristic parameters include at least one of node identification information, node processing time, and node load; An analysis is performed based on the clustering results of all nodes to determine the target approval node to be optimized, and an optimization strategy is executed for the target approval node.

4. The method according to claim 3, characterized in that: The step of analyzing and processing all node clustering results to determine the target approval node to be optimized, and executing the optimization strategy for the target approval node includes: Selecting a first clustering result that meets the labor efficiency optimization condition from all node clustering results, and executing a labor efficiency optimization strategy for the target approval node corresponding to the first clustering result; and / or, At least two second clustering results meeting the node merging condition are selected from all node clustering results, and merging control is performed on the approval nodes respectively corresponding to the at least two second clustering results.

5. The method according to claim 4, characterized in that The human efficiency optimization conditions include at least one of the following: The node processing time corresponding to all or part of the node information in the first clustering result is greater than a preset time; The node loads corresponding to all or part of the node information in the first clustering result are all greater than a load threshold.

6. The method according to claim 5, characterized in that The labor efficiency optimization strategy includes at least one of the following: For the approval node corresponding to the first clustering result, the approval personnel configuration control corresponding to the senior functional personnel is performed to replace the original approval personnel in the approval node corresponding to the first clustering result with the senior functional personnel; For the approval node corresponding to the first clustering result, personnel addition control is performed to add new functional personnel based on the original approval personnel in the approval node corresponding to the first clustering result; The personnel information of the approving personnel is obtained according to the approval node corresponding to the first clustering result, and personnel training control is performed according to the personnel information.

7. The method according to claim 6, characterized in that The obtaining of personnel information of the approving personnel according to the approval node corresponding to the first clustering result, and performing personnel training control according to the personnel information, includes: Acquiring personnel information of the approving personnel according to the approval node corresponding to the first clustering result; Determine the number of training sessions based on the professional experience in the personnel information; Generate a training task according to the review operation guidance information associated with the approval node corresponding to the first clustering result and the number of training times; According to the notification receiving identifier in the personnel information, the training task is sent to the user terminal corresponding to the notification receiving identifier.

8. The method according to claim 4, characterized in that The node information also includes node sequence information, and the characteristic parameters also include node approval results; The node merging condition includes at least one of the following: The node identification information respectively corresponding to the at least two second clustering results are different from each other; According to the node sequence information respectively corresponding to the at least two second clustering results, determining whether the approval nodes respectively corresponding to the at least two clustering results have a sequence on the approval path; The node approval results corresponding to the at least two second clustering results are the same.

9. A computing device, characterized in that include: A processor and a memory storing a computer program, when the processor runs the computer program, implements the steps of the control method of the approval process according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: A computer program is stored, and when the computer program is executed by a processor, the steps of the control method of the approval process described in any one of claims 1 to 8 are implemented.

Images