Process management method, device, equipment, storage medium, computer program product

Abstract

The present application discloses a process management method, apparatus, device, storage medium, and computer program product, which are used to solve the problems that the existing process management solutions have low efficiency and cannot perform real-time correction of processes according to actual business requirements. The method includes: receiving a process correction work order input by a user through a pre-trained work order portrait template, and determining the correction features corresponding to the process correction work order; according to the correction features, using a pre-trained risk identification model to perform risk identification on the process correction work order, and determining the risk level corresponding to the process correction work order; according to the risk level, determining the approval node corresponding to the process correction work order, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; and generating an approval process corresponding to the process correction work order according to the optimal approval path.

Description

Technical Field

[0001] This application relates to the field of computer information technology, and particularly to a process management method, apparatus, device, storage medium, and computer program product. Background Art

[0002] With the expansion of enterprise scale and the increase in business complexity, enterprises' demand for process management is growing day by day. The traditional manual-dependent process management method has become difficult to meet the needs of modern enterprises. In the process of modern enterprise management and project R & D, the automation and intelligence of management processes have become the key to improving efficiency and reducing costs.

[0003] Currently, as an important tool for managing, automating, and optimizing enterprise processes, the workflow engine is increasingly widely used. However, in actual applications, existing process management solutions often rely on pre-manually defined process nodes and often face many challenges in terms of process modification and optimization. For example, in existing process automation management solutions, when it is necessary to modify and correct data for a designed process according to actual business needs, it often needs to be manually operated, which is time-consuming, less accurate, and inefficient. On the other hand, existing process correction solutions are usually carried out regularly and cannot achieve real-time monitoring and correction of business processes, resulting in data that may be outdated when used, affecting normal business operations. In summary, existing process management solutions have been unable to adapt to the rapid changes in enterprise business needs.

[0004] Therefore, how to accurately and efficiently implement the modification and management of business processes according to the actual changes in business needs has become an urgent problem to be solved currently. Summary of the Invention

[0005] An embodiment of this application provides a process management method to solve the problems that existing process management solutions are less efficient and cannot perform real-time process correction according to actual business needs.

[0006] An embodiment of this application also provides a process management apparatus to solve the problems that existing process management solutions are less efficient and cannot perform real-time process correction according to actual business needs.

[0007] An embodiment of this application also provides a process management device to solve the problems that existing process management solutions are less efficient and cannot perform real-time process correction according to actual business needs.

[0008] An embodiment of this application also provides a computer-readable storage medium to solve the problems that existing process management solutions are less efficient and cannot perform real-time process correction according to actual business needs.

[0009] A computer program product is used to solve the problems that the existing process management solutions have low efficiency and cannot perform real-time correction of processes according to actual business requirements.

[0010] The embodiments of this application adopt the following technical solutions:

[0011] A process management method includes: receiving a process correction work order input by a user through a pre-trained work order portrait template, and determining the correction features corresponding to the process correction work order, where the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, using a pre-trained risk identification model to perform risk identification on the process correction work order, and determining the risk level corresponding to the process correction work order; according to the risk level, determining the approval node corresponding to the process correction work order, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; and generating an approval process corresponding to the process correction work order according to the optimal approval path.

[0012] A process management device includes: a correction work order receiving unit for receiving a process correction work order input by a user through a pre-trained work order portrait template, and determining the correction features corresponding to the process correction work order, where the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; a risk identification unit for performing risk identification on the process correction work order according to the correction features by using a pre-trained risk identification model, and determining the risk level corresponding to the process correction work order; an approval path generation unit for determining the approval node corresponding to the process correction work order according to the risk level, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; and a process generation unit for generating an approval process corresponding to the process correction work order according to the optimal approval path.

[0013] A process management device includes:

[0014] A processor; and a memory arranged to store computer-executable instructions, which when executed cause the processor to perform the following operations: receiving a process correction work order input by a user through a pre-trained work order profile template, determining correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, using a pre-trained risk identification model to perform risk identification on the process correction work order, determining the risk level corresponding to the process correction work order; according to the risk level, determining the approval node corresponding to the process correction work order, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; according to the optimal approval path, generating an approval process corresponding to the process correction work order.

[0015] A computer-readable storage medium stores one or more programs, which when executed by an electronic device including multiple application programs, cause the electronic device to perform the following operations: receiving a process correction work order input by a user through a pre-trained work order profile template, determining correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, using a pre-trained risk identification model to perform risk identification on the process correction work order, determining the risk level corresponding to the process correction work order; according to the risk level, determining the approval node corresponding to the process correction work order, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; according to the optimal approval path, generating an approval process corresponding to the process correction work order.

[0016] A computer program product includes a computer program, which when executed by a processor implements: receiving a process correction work order input by a user through a pre-trained work order profile template, determining correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, using a pre-trained risk identification model to perform risk identification on the process correction work order, determining the risk level corresponding to the process correction work order; according to the risk level, determining the approval node corresponding to the process correction work order, and based on the approval node, determining the optimal approval path corresponding to the process correction work order; according to the optimal approval path, generating an approval process corresponding to the process correction work order.

[0017] The above at least one technical solution adopted in the embodiments of the present application can achieve the following beneficial effects:

[0018] By using the process management method provided in the embodiments of the present application, when a user needs to initiate a process correction work order, the correction work order parameters input by the user can be converted into a general process correction work order through a pre-established work order portrait template. Furthermore, when the system receives the process correction work order input by the user through the pre-trained work order portrait template, it can determine the correction features corresponding to the process correction work order, including: the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order. Then, the system can use the pre-trained risk identification model to identify the risk of the process correction work order based on the correction features, determine the risk level corresponding to the process correction work order, and based on the risk level, determine the approval node corresponding to the process correction work order, and determine the optimal approval path corresponding to the process correction work order based on the approval node. According to the optimal approval path, an approval process corresponding to the process correction work order is generated. By using the process management method provided in the embodiments of the present application, on the one hand, different format work orders submitted by different branch companies or different departments can be automatically and efficiently converted into a general format through a pre-established work order portrait template. Subsequently, the system can determine the approval node and approval department corresponding to the work order based on these general format process correction work orders, greatly improving the generation efficiency of subsequent cross-department approval processes. On the other hand, when generating the approval process in the embodiments of the present application, the risk level corresponding to the work order will be determined first, and the approval process for each correction work order will be dynamically generated under the drive of risk according to the risk level corresponding to the work order, which not only avoids redundant approval processes, but also ensures that the approval process covers the highest risk verification requirements, greatly improving the security and compliance of process management and the processing efficiency of process work orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The schematic embodiments and descriptions thereof of the present application are used to explain the present application, and do not constitute an improper limitation to the present application. In the drawings:

[0020] Figure 1 is a specific process schematic diagram of a process management method provided by an embodiment of the present application;

[0021] Figure 2 is a specific structural schematic diagram of a work order portrait template provided by an embodiment of the present application;

[0022] Figure 3 is a specific structural schematic diagram of a process management device provided by an embodiment of the present application;

[0023] Figure 4 This is a schematic diagram of the specific structure of a process management device provided by an embodiment of the present application. Detailed implementation manners

[0024] To make the objectives, technical solutions, and advantages of the present application clearer, the technical solutions of the present application will be clearly and completely described below in conjunction with specific embodiments of the present application and the corresponding drawings. Apparently, the described embodiments are only a part rather than all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application.

[0025] To solve the problems that the existing process management solutions are inefficient and cannot perform real-time correction of processes according to actual business requirements, an embodiment of the present application provides a process management system and a process management method based on the process management system.

[0026] The execution subject of the process management method provided by the embodiment of the present application may be, but is not limited to, at least one of a process management server, a work order management server, a process approval server, etc.; or, the execution subject of this method may also be a process management system running on a server; in addition, the execution subject of this method may also be a terminal device used by a user or a process management personnel, or even a process management application installed on the terminal device. For the sake of convenience of description, the embodiments of the present invention are all described by taking the execution subject as a process management system running on a server as an example. It can be understood that taking the execution subject of this method as a process management system on a server is only an exemplary description and should not be construed as a limitation on this method.

[0027] The specific implementation process diagram of the process management method provided by the present application is as Figure 1 shown, and mainly includes the following steps:

[0028] Step 11, receiving a process correction work order input by a user through a pre-trained work order portrait template, and determining the correction features corresponding to the process correction work order;

[0029] In one implementation manner, the process management system may construct a work order portrait template according to the following sub-steps, which may specifically include:

[0030] Sub-step 1101, extracting features from each historical correction work order in the obtained historical correction work order set;

[0031] Specifically, in the embodiment of the present application, the process management system may extract fields from the work order logs corresponding to the historical correction work orders, and then obtain the features of the historical correction work orders.

[0032] It should be noted here that the features extracted by the process management system from the work order logs corresponding to the historical correction work orders can include the following types:

[0033] a. Text features corresponding to the historical correction work orders;

[0034] Among them, the text features correspond to the basic information of the historical correction work orders, which can specifically include the work order description and the correction operation log corresponding to the historical correction work orders.

[0035] b. Category features corresponding to the historical correction work orders;

[0036] Among them, the category features can include the work order type corresponding to the historical correction work order (such as auto insurance claims, insurance application, or non-auto insurance claims, etc.), the work order submission time, and the corresponding processing system of the work order.

[0037] c. Historical risk labels corresponding to the historical correction work orders;

[0038] Among them, the historical risk label is the actual risk level marked by the process management system for these historical correction work orders (for example, levels 1-5).

[0039] In addition, the process management system can also extract technical features corresponding to these historical correction work orders through the work order logs, such as: the length of the correction script corresponding to the historical correction work order, the amount of corrected data, and the amount of corrected data for the corresponding key data tables, etc. Among them, the length of the correction script is used to represent the complexity of the correction script corresponding to the historical correction work order, which can specifically include the number of code lines of the correction script, the number of logical branches (such as the nesting level of if), the depth of function calls, etc.; the amount of corrected data is used to represent the number of data tables affected by the correction work order and the total number of data rows involved (such as the number of rows affected by the UPDATE statement); the key data table refers to the data table that is pre-set to involve user key data. For example, the data table involving fields such as amount, password, and user permissions, then the amount of corrected data for the corresponding key data table refers to whether the correction work order involves modifying the data in these key data tables, and the number of modifications to the data in the key data tables.

[0040] After completing the extraction of the above features, the process management system can perform preprocessing such as data cleaning, missing value processing, and format conversion on the extracted features. After completing the preprocessing, the process management system can perform feature engineering processing on these features, specifically including:

[0041] (1)Extract key features for text features. Specifically, in the embodiments of this application, the keyword vectors of work order texts can be extracted using, for example, term frequency–inverse document frequency (TF-IDF) or pre-trained language models (Bidirectional Encoder Representations from Transformers, BERT).

[0042] (2)Perform encoding processing for category features: Specifically, for the work order type feature, one-hot encoding can be performed.

[0043] (3)Standardize the risk labels: Specifically, in the embodiments of this application, numerical risk labels can be normalized (for example, using the Z-Score algorithm for normalization).

[0044] After the above processing, standardized features are finally output, and based on these standardized features, structured input is provided for training the work order portrait template.

[0045] Sub-step 1102: Cluster the historical corrected work order set according to the text features, category features, and historical risk labels obtained by executing sub-step 1101, and obtain historical corrected work order subsets corresponding to multiple work order types;

[0046] Specifically, in the embodiments of this application, the process management system can use the K-means or DBSCAN algorithm to cluster the standardized features obtained by executing the above sub-step 1101, identify similar work order groups, and obtain historical corrected work order subsets corresponding to multiple work order types.

[0047] Sub-step 1103: Determine the correction scripts, correction fields, and constraint conditions corresponding to each work order type according to the historical corrected work order subsets respectively;

[0048] Specifically, in the embodiments of this application, for each type of work order (such as auto insurance work orders, non-auto insurance work orders, etc.), the process management system can extract its high-frequency operation steps (such as SQL correction statement patterns), co-occurring fields (such as the required field order_id), operation constraint conditions (such as only allowing DML statements), and other fields. Subsequently, the most frequent fields, etc., can be used as the underlying configuration to generate work order portrait templates corresponding to each work order type.

[0049] Sub-step 1104: Generate work order portrait templates corresponding to each work order type according to the correction scripts, correction fields, and constraint conditions corresponding to each work order type respectively.

[0050] The process management system can generate a work order portrait template corresponding to each work order type with the most frequent fields obtained by executing sub-step 1103 as the underlying configuration.

[0051] By executing the above sub-steps, the process management system can pre-construct a work order portrait template corresponding to various work orders. The structure of the work order portrait template is as Figure 2 shown. Based on this work order portrait template, the user can use the work order correction content as the basic input of the work order portrait template. Each work order portrait template base includes a portrait name (i.e., the work order name, such as: auto insurance work order portrait template, non-auto insurance work order portrait template, claim work order portrait template, and underwriting work order portrait template, etc.), a portrait description (work order description), the systems involved, and portrait input parameters 1, input parameter 2, input parameter 3.. input parameter N, etc. The user only needs to input the corresponding parameters into the corresponding modules of the work order portrait template according to the classification, and the work order portrait template can automatically generate the correction script corresponding to the process correction work order.

[0052] Step 12, according to the correction features obtained by executing Step 11, use the pre-trained risk identification model to identify the risks of the process correction work order and determine the risk level corresponding to the process correction work order;

[0053] In the embodiment of the present application, the process management system can construct a risk identification model based on a model-agnostic post hoc explanation method (Local Interpretable Model-Agnostic Explanation, LIME).

[0054] Specifically, in the embodiment of the present application, the process management system can train the risk identification model according to the following sub-steps, including:

[0055] Sub-step 1201, determine multiple risk factors according to the correction script length, the amount of correction data, and the amount of correction data of the key data table corresponding to the process correction work order;

[0056] In the embodiment of the present application, the process management system can determine the correction script length, the amount of correction data, and the amount of correction data of the key data table corresponding to the process correction work order as three core risk factors according to the business rules.

[0057] Sub-step 1202, according to the risk factors determined by executing sub-step 1201, respectively obtain the risk characteristic values corresponding to each risk factor in each historical correction work order;

[0058] Specifically, the process management system can extract the length of the correction script, the amount of corrected data for each historical correction work order, and the amount of corrected data for the key data tables corresponding to the process correction work order from the work order log corresponding to the historical correction work order, and then determine the risk characteristic value corresponding to each historical correction work order according to the manual review result. In the embodiment of the present application, it can be assumed that the risk characteristic value R ∈ level 1-5, where 5 is the highest risk.

[0059] Sub-step 1203, normalize the risk characteristic value obtained by executing sub-step 1202 to obtain a standardized risk characteristic value, and construct a training data set according to the standardized risk characteristic value;

[0060] In the embodiment of the present application, the process management system can normalize the length of the correction script, the amount of corrected data, and the amount of corrected data for the key data tables corresponding to each historical correction work order (such as Z-Score) to eliminate the dimension difference. At the same time, for the data with unbalanced risk characteristic values, the Synthetic Minority Oversampling Technique (SMOTE) can be used for processing, so as to avoid the model being biased towards the majority class, and then construct a training data set with the standardized risk characteristic value obtained after the above processing.

[0061] Sub-step 1204, input the standardized risk characteristic value in the training data set obtained by executing sub-step 1203 into the initial risk identification model, and obtain the risk prediction result corresponding to each historical correction work order;

[0062] In one implementation, the process management system can construct an initial risk identification model based on the Distributed Gradient Boosting Library (XGBoost). By inputting the standardized risk characteristic value into the initial risk identification model, the risk prediction result output by the initial risk identification model can be obtained.

[0063] Sub-step 1205, sample and generate a perturbation sample set according to the risk characteristic value corresponding to any historical correction work order;

[0064] It should be noted here that since the above initial risk identification model belongs to a complex black-box model, there is a huge gap between its prediction ability and the logic understandable by humans. In order to enhance the transparency and trust of the risk identification model, in the embodiments of this application, the process management system can, by constructing an interpretable risk prediction model (Local Interpretable Model-agnostic Explanations, LIME), provide an intuitive explanation of the individual prediction results of the complex model (i.e., the risk identification model) based on the LIME model, helping users understand "why the model gives a specific prediction result for a certain sample", so as to achieve the user's understanding of the prediction logic of the risk identification model, and further improve the transparency and trust of the model.

[0065] Specifically, in the embodiments of this application, the process management system can first generate N perturbed samples by randomly sampling centered on the current historical correction work order to be explained.

[0066] For example, for a single historical correction work order to be explained (assuming that the risk characteristic value R of this historical correction work order is 4), taking the characteristics of this historical correction work order (such as the length of the correction script = 6, the amount of corrected data = 500 lines, the amount of corrected data in the key data table = 3) as the center, delimit the neighborhood range of local interpretation, and by randomly perturbing the characteristics of this historical correction work order (such as randomly increasing or decreasing the number of code lines, adjusting the affected lines, replacing key fields), generate a "virtual work order" data set that is similar but slightly different from this historical correction work order as the perturbed sample set. For example, change the length of the correction script of this historical correction work order from 6 to 4 randomly while keeping other characteristics unchanged, and generate multiple simulated work orders.

[0067] Sub-step 1206: Train a local interpretation proxy model according to the perturbed sample set obtained by executing sub-step 1205;

[0068] Sub-step 1207: Interpret the risk prediction result output by the initial risk identification model according to the local interpretation proxy model obtained by executing sub-step 1206, and determine the contribution degree weights of each risk factor to the prediction result of a single historical correction work order;

[0069] In the embodiments of this application, the process management system can use linear regression or logistic regression algorithms to train the proxy model. The task of the trained proxy model is to imitate the prediction results of the initial risk identification model for the perturbed data. After training, the coefficients of the proxy model can reflect the positive / negative influence degrees of each risk factor on the current prediction result, that is, the contribution degree weights of each risk factor to the prediction result of a single historical correction work order. The larger the absolute value of the contribution degree weight, the more significant the influence of this risk factor on the current prediction result.

[0070] Sub-step 1208: Iteratively optimize the initial risk identification model according to the contribution degree weights obtained by executing sub-step 1209 to obtain the trained risk identification model.

[0071] In one implementation, the process management system can convert the contribution degree weights output by the local interpretation agent model into operable business rules, and then continuously optimize and iterate the risk identification model and interpretation rules according to the business rules and actual application feedback.

[0072] Furthermore, the process management system can use the trained risk identification model to identify risks for the process correction work order and determine the risk level corresponding to the process correction work order.

[0073] Step 13: Determine the approval node corresponding to the process correction work order according to the risk level obtained by executing Step 12, and based on the approval node, determine the optimal approval path corresponding to the process correction work order;

[0074] In one implementation, the process management system can first determine the approval node corresponding to the process correction work order according to the determined risk level. Specifically, the process management system can match the set of approval departments (i.e., the corresponding approval node) required for the correction work order of this risk level through the risk level hit rule engine, and then determine the hierarchical relationship between the approval nodes according to the approval authority of the approval departments, and generate an approval authority binary tree corresponding to the approval node according to the hierarchical relationship.

[0075] For example, the process management system can determine the approval authority and hierarchical relationship corresponding to each approval department according to the multi-level department organizational structure of the company (such as head office - business department - business branch - business group), and based on this, convert the company organizational structure into the corresponding approval authority binary tree. For example, for the approval departments at the same level, the approval departments at the same level can be sorted according to the approval authority (such as risk control ability), and the approval department with higher approval authority can be used as the left child node, and the lower-level departments are recursively constructed as child nodes.

[0076] After completing the construction of the approval authority binary tree, the process management system can further determine the risk level corresponding to each approval node. For example, if the risk level R corresponding to the approval node (Claims Department 1) is 3, it means that the minimum risk level for triggering the approval of this approval node (Claims Department 1) is level 3.

[0077] When it is determined according to the generated approval authority binary tree that the current process correction work order triggers cross-department approval of multiple departments, the process management system can find the common approval level from the paths of the approval nodes corresponding to these approval departments, that is, find the lowest common ancestor node (LCA).

[0078] Specifically, the process management system can traverse the paths from each approval department to the root node (such as the head office) based on the approval authority binary tree, and determine the common superior node with the highest level in all paths as the common ancestor node.

[0079] Determine the node paths from each approval node to the least common ancestor node, and based on the path depth algorithm, determine the path depths of each node path. Specifically, the process management system can start from the LCA node, traverse all sub-paths downward, and select the path with the maximum depth as the approval path corresponding to the process correction work order.

[0080] At the same time, it should be noted here that if there are multiple approval paths with the same path depth, the path where the approval node with a higher weight (such as the left child node) is located can be preferentially retained as the approval path. Finally, merge the repeated approval nodes (i.e., approval departments) in the upper and lower paths to generate the optimal approval path.

[0081] By adopting the above-mentioned approval path generation method based on the least common ancestor and the maximum depth of the binary tree provided by the embodiments of the present application, when approving cross-department work orders, it can be automatically merged to the common superior (such as the head office), reducing the repeated approval levels. At the same time, for the risk levels corresponding to different work orders, the process management system can automatically trigger different approval paths according to the risk levels. For example, for a process correction work order with a high risk level (such as R = 5), the longest path (the strictest review chain) is forcibly triggered, and for a low-risk process correction work order (R = 2), only a shallow approval is triggered, realizing the dynamic optimization of the sensitivity of process approval. The process management system can adaptively generate an approval process that takes into account both efficiency and risk control, and at the same time supports the flexible expansion of complex organizational structures.

[0082] Step 14, generate the approval process corresponding to the process correction work order according to the optimal approval path determined by executing Step 13.

[0083] Using the process management method provided by the embodiments of the present application, when a user needs to initiate a process correction work order, the correction work order parameters input by the user can be converted into a general process correction work order through a pre-established work order portrait template. Then, when the system receives the process correction work order input by the user through the pre-trained work order portrait template, the correction features corresponding to the process correction work order can be determined, including: the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order. Then, the system can use the pre-trained risk identification model to identify the risk of the process correction work order according to the correction features, determine the risk level corresponding to the process correction work order, and determine the approval node corresponding to the process correction work order according to the risk level. Based on the approval node, determine the optimal approval path corresponding to the process correction work order, and generate the approval process corresponding to the process correction work order according to the optimal approval path. Using the process management method provided by the embodiments of the present application, on the one hand, different format work orders submitted by different branch companies or different departments can be automatically and efficiently converted into a general format through a pre-established work order portrait template. Subsequently, the system can determine the approval node and approval department corresponding to the work order based on these general format process correction work orders, greatly improving the generation efficiency of subsequent cross-department approval processes. On the other hand, when generating the approval process, the embodiments of the present application will first determine the risk level corresponding to the work order, and dynamically generate the approval process for each correction work order under the drive of risk according to the risk level corresponding to the work order, which not only avoids redundant approval processes, but also ensures that the approval process covers the highest risk verification requirements, greatly improving the security and compliance of process management and the processing efficiency of process work orders.

[0084] In one implementation, the embodiments of the present application further provide a process management device to solve the problems of low efficiency in existing process management solutions and inability to perform real-time process correction according to actual business requirements. The specific structural schematic diagram of the process management device is as Figure 3 shown, including: a correction work order receiving unit 31, a risk identification unit 32, an approval path generation unit 33, and a process generation unit 34.

[0085] Among them, the correction work order receiving unit 31 is used to receive the process correction work order input by the user through the pre-trained work order portrait template, and determine the correction features corresponding to the process correction work order. Among them, the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order.

[0086] A risk identification unit 32, configured to perform risk identification on the process correction work order according to the correction feature by using a pre-trained risk identification model, and determine the risk level corresponding to the process correction work order;

[0087] An approval path generation unit 33, configured to determine the approval node corresponding to the process correction work order according to the risk level, and determine the optimal approval path corresponding to the process correction work order based on the approval node;

[0088] A process generation unit 34, configured to generate an approval process corresponding to the process correction work order according to the optimal approval path.

[0089] In an implementation manner, it further includes a work order portrait template training unit, which is specifically configured to: extract features from each historical correction work order in the obtained historical correction work order set to obtain the text features, category features, and historical risk labels corresponding to the historical correction work order; wherein, the text features include the work order description and the correction operation log corresponding to the historical correction work order, and the category features include the work order type and the system corresponding to the historical correction work order; cluster the historical correction work order set according to the text features, the category features, and the historical risk labels to obtain historical correction work order subsets corresponding to multiple work order types; respectively determine the correction scripts, correction fields, and constraint conditions corresponding to each work order type according to the historical correction work order subsets; respectively generate work order portrait templates corresponding to each work order type according to the correction scripts, correction fields, and constraint conditions corresponding to each work order type.

[0090] In an implementation manner, it further includes a risk identification model training unit, which is specifically configured to: determine multiple risk factors according to the correction script length, the correction data volume, and the correction data volume of the key data table corresponding to the process correction work order; respectively obtain the risk feature values corresponding to each risk factor in each historical correction work order according to the risk factors; perform normalization processing on the risk feature values to obtain standardized risk feature values, and construct a training data set according to the standardized risk feature values; input the standardized risk feature values in the training data set into an initial risk identification model to obtain risk prediction results corresponding to each historical correction work order; sample and generate a perturbation sample set according to the risk feature values corresponding to any historical correction work order; train a local interpretation proxy model according to the perturbation sample set; interpret the risk prediction results output by the initial risk identification model according to the local interpretation proxy model to determine the contribution degree weights of each risk factor to the prediction result of a single historical correction work order; perform iterative optimization on the initial risk identification model according to the contribution degree weights to obtain a trained risk identification model.

[0091] In one embodiment, the approval path generation unit 33 is specifically configured to: respectively determine the approval departments corresponding to each approval node; determine the approval authorities corresponding to each approval node according to the approval departments; determine the hierarchical relationship between each of the approval nodes according to the approval authorities, and generate an approval authority binary tree corresponding to the approval nodes according to the hierarchical relationship; and determine the optimal approval path corresponding to the process correction work order according to the approval authority binary tree.

[0092] In one embodiment, the approval path generation unit 33 is specifically configured to: determine the least common ancestor node corresponding to each of the approval departments according to the approval authority binary tree; determine the node paths from each of the approval nodes to the least common ancestor node, and determine the path depths of each of the node paths based on the path depth algorithm; and determine the node path with the maximum path depth as the optimal approval path corresponding to the process correction work order.

[0093] When using the process management device provided in the embodiments of the present application, when a user needs to initiate a process correction work order, the correction work order parameters input by the user can be converted into a general process correction work order through a pre-established work order portrait template. Furthermore, when the system receives the process correction work order input by the user through the pre-trained work order portrait template, the correction features corresponding to the process correction work order can be determined, including: the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order. Furthermore, the system can use the pre-trained risk identification model to identify the risk of the process correction work order according to the correction features, and determine the risk level corresponding to the process correction work order; and determine the approval node corresponding to the process correction work order according to the risk level, and determine the optimal approval path corresponding to the process correction work order based on the approval node; and generate an approval process corresponding to the process correction work order according to the optimal approval path. When using the process management method provided in the embodiments of the present application, on the one hand, different format work orders submitted by different branch companies or different departments can be automatically and efficiently converted into a general format through the pre-established work order portrait template. Subsequently, the system can determine the approval node and approval department corresponding to the work order based on these general format process correction work orders, greatly improving the generation efficiency of the subsequent cross-department approval process. On the other hand, when generating the approval process in the embodiments of the present application, the risk level corresponding to the work order will be determined first, and an approval process for each correction work order will be dynamically generated under the drive of risk according to the risk level corresponding to the work order, which not only avoids redundant approval processes, but also ensures that the approval process covers the highest risk verification requirements, greatly improving the security and compliance of process management and the processing efficiency of process work orders.

[0094] Figure 4 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Please refer toFigure 4 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. Among them, the memory may include internal memory, such as high-speed random access memory (RAM), and may also include non-volatile memory, such as at least one disk memory, etc. Of course, the electronic device may also include other hardware required for other services.

[0095] The processor, network interface, and memory can be interconnected through an internal bus, and the internal bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 4 only a bidirectional arrow is used in the figure, but it does not mean that there is only one bus or one type of bus.

[0096] The memory is used to store programs. Specifically, the program may include program code, and the program code includes computer operation instructions. The memory can include internal memory and non-volatile memory, and provide instructions and data to the processor.

[0097] The processor reads the corresponding computer program from the non-volatile memory into the internal memory and then runs it, forming a process management device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the following operations:

[0098] Receive a process correction work order input by the user through a pre-trained work order portrait template, and determine the correction features corresponding to the process correction work order, where the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, use a pre-trained risk identification model to perform risk identification on the process correction work order, and determine the risk level corresponding to the process correction work order; according to the risk level, determine the approval node corresponding to the process correction work order, and based on the approval node, determine the optimal approval path corresponding to the process correction work order; according to the optimal approval path, generate the approval process corresponding to the process correction work order.

[0099] The above is as described in this application Figure 4The method executed by the process management electronic device disclosed in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor or instructions in the form of software. The above processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed and completed by a hardware decoding processor, or executed and completed by a combination of hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, flash memory, read-only memory, programmable read-only memory, or electrically erasable programmable memory, register, etc. This storage medium is located in the memory, and the processor reads the information in the memory and combines its hardware to complete the steps of the above method.

[0100] Of course, in addition to the software implementation, the electronic device of the present application does not exclude other implementation manners, such as a logic device or a combination of software and hardware, etc. That is to say, the execution subject of the following processing flow is not limited to each logic unit, and may also be hardware or a logic device.

[0101] The embodiments of the present application also propose a computer-readable storage medium that stores one or more programs. The one or more programs include instructions that, when executed by a portable electronic device including multiple application programs, can enable the portable electronic device to execute Figure 1 the method of the illustrated embodiment and specifically used to perform the following operations:

[0102] Receive a process correction work order input by the user through a pre-trained work order portrait template, and determine the correction features corresponding to the process correction work order, where the correction features include the length of the correction script corresponding to the process correction work order, the amount of correction data corresponding to the process correction work order, and the amount of correction data of the key data table corresponding to the process correction work order; according to the correction features, use a pre-trained risk identification model to identify the risk of the process correction work order, and determine the risk level corresponding to the process correction work order; according to the risk level, determine the approval node corresponding to the process correction work order, and based on the approval node, determine the optimal approval path corresponding to the process correction work order; according to the optimal approval path, generate the approval process corresponding to the process correction work order.

[0103] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0104] The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0105] These computer program instructions can also be stored in a computer-readable memory that can guide a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured product including an instruction device, and the instruction device realizes the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0106] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one process Figure 1 one process or multiple processes and / or blocks Figure 1 steps for the functions specified in one block or multiple blocks.

[0107] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0108] The memory may include non-permanent memory in the computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of, for example, read-only memory (ROM) or flash memory (flash RAM). The memory is an example of a computer-readable medium.

[0109] Computer-readable media includes permanent and non-permanent, removable and non-removable media and can store information by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0110] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0111] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0112] The above are only the embodiments of the present application and are not used to limit the present application. For those skilled in the art, various changes and modifications can be made to the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. A process management method, characterized in that: include: Perform feature extraction on each historical correction work order in the acquired historical correction work order set to obtain text features, category features and historical risk labels corresponding to the historical correction work order; wherein, the text features include the work order description and correction operation log corresponding to the historical correction work order, and the category features include the work order type and system corresponding to the historical correction work order; cluster the historical correction work order set based on the text features, the category features and the historical risk labels to obtain historical correction work order subsets corresponding to multiple work order types; determine the correction script, correction field and constraint conditions corresponding to each work order type based on the historical correction work order subsets; generate a work order portrait template corresponding to each work order type based on the correction script, correction field and constraint conditions corresponding to each work order type; Receive a process correction work order input by a user through the work order portrait template, and determine correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the number of data tables affected by the process correction work order and the total number of data rows involved, and the number of data modified in the key data table corresponding to the process correction work order; Determine multiple risk factors based on the length of the correction script, the number of data tables affected by the process correction work order, the total number of data rows involved in the process correction work order, and the number of data modifications in the key data table corresponding to the process correction work order; obtain the risk characteristic values corresponding to each risk factor in each historical correction work order based on the risk factors; normalize the risk characteristic values to obtain standardized risk characteristic values, and construct a training data set based on the standardized risk characteristic values; input the standardized risk characteristic values in the training data set into the initial risk identification model to obtain the risk prediction results corresponding to each historical correction work order; sample and generate a perturbation sample set based on the risk characteristic value corresponding to any historical correction work order; train a local explanation proxy model based on the perturbation sample set; interpret the risk prediction results output by the initial risk identification model based on the local explanation proxy model, and determine the contribution weight of each risk factor to the prediction result of a single historical correction work order; iteratively optimize the initial risk identification model based on the contribution weight to obtain a trained risk identification model; Based on the correction features, using the risk identification model, performing risk identification on the process correction work order, and determining the risk level corresponding to the process correction work order; Determining, according to the risk level, an approval node corresponding to the process correction work order, and determining, based on the approval node, an optimal approval path corresponding to the process correction work order; An approval process corresponding to the process correction work order is generated according to the optimal approval path.

2. The method according to claim 1, characterized in that Determining the optimal approval path corresponding to the process modification work order based on the approval node specifically includes: Determine the approval department corresponding to each approval node; Determine the approval authority corresponding to each approval node according to the approval department; According to the approval authority, determining the hierarchical relationship between the approval nodes, and generating a binary tree of approval authority corresponding to the approval nodes according to the hierarchical relationship; According to the approval authority binary tree, the optimal approval path corresponding to the process correction work order is determined.

3. The method according to claim 2, characterized in that Determining the optimal approval path corresponding to the process modification work order based on the approval authority binary tree specifically includes: Determine the least common ancestor node corresponding to each of the approval departments according to the approval authority binary tree; Determining a node path from each of the approval nodes to the least common ancestor node, and determining a path depth of each of the node paths based on a path depth algorithm; The node path with the largest path depth is determined as the optimal approval path corresponding to the process correction work order.

4. A process management device, characterized in that: include: A work order portrait template generation unit is used to extract features from each historical correction work order in the acquired historical correction work order set to obtain text features, category features and historical risk labels corresponding to the historical correction work order; wherein, the text features include the work order description and correction operation log corresponding to the historical correction work order, and the category features include the work order type and system corresponding to the historical correction work order; according to the text features, the category features and the historical risk labels, the historical correction work order set is clustered to obtain a subset of historical correction work orders corresponding to multiple work order types; according to the historical correction work order subsets, the correction script, correction field and constraint conditions corresponding to each work order type are determined; according to the correction script, correction field and constraint conditions corresponding to each work order type, a work order portrait template corresponding to each work order type is generated; A correction work order receiving unit is configured to receive a process correction work order input by a user using a pre-trained work order portrait template, and determine correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the number of data tables affected by the process correction work order, the total number of data rows involved in the process correction work order, and the number of data modifications in the key data table corresponding to the process correction work order; A risk identification model training unit is used to determine multiple risk factors based on the length of the correction script, the number of data tables affected by the process correction work order, the total number of data rows involved in the process correction work order, and the number of data modifications in the key data table corresponding to the process correction work order; based on the risk factors, respectively obtain the risk characteristic values corresponding to each risk factor in each historical correction work order; normalize the risk characteristic values to obtain standardized risk characteristic values, and construct a training data set based on the standardized risk characteristic values; input the standardized risk characteristic values in the training data set into the initial risk identification model to obtain the risk prediction results corresponding to each historical correction work order; sample and generate a perturbation sample set based on the risk characteristic value corresponding to any historical correction work order; train a local explanation proxy model based on the perturbation sample set; interpret the risk prediction results output by the initial risk identification model based on the local explanation proxy model, and determine the contribution weight of each risk factor to the prediction result of a single historical correction work order; iteratively optimize the initial risk identification model based on the contribution weight to obtain a trained risk identification model; a risk identification unit, configured to perform risk identification on the process correction work order based on the correction feature and using a pre-trained risk identification model, and determine a risk level corresponding to the process correction work order; An approval path generating unit, configured to determine an approval node corresponding to the process modification work order according to the risk level, and determine an optimal approval path corresponding to the process modification work order based on the approval node; The process generation unit is used to generate the approval process corresponding to the process correction work order according to the optimal approval path.

5. The device according to claim 4, characterized in that The approval path generation unit is specifically used to: Determine the approval department corresponding to each approval node; Determine the approval authority corresponding to each approval node according to the approval department; According to the approval authority, determining the hierarchical relationship between the approval nodes, and generating a binary tree of approval authority corresponding to the approval nodes according to the hierarchical relationship; According to the approval authority binary tree, the optimal approval path corresponding to the process correction work order is determined.

6. A process management device, characterized in that: include: processor; as well as a memory arranged to store computer-executable instructions which, when executed, cause the processor to: Perform feature extraction on each historical correction work order in the acquired historical correction work order set to obtain text features, category features and historical risk labels corresponding to the historical correction work order; wherein, the text features include the work order description and correction operation log corresponding to the historical correction work order, and the category features include the work order type and system corresponding to the historical correction work order; cluster the historical correction work order set based on the text features, the category features and the historical risk labels to obtain historical correction work order subsets corresponding to multiple work order types; determine the correction script, correction field and constraint conditions corresponding to each work order type based on the historical correction work order subsets; generate a work order portrait template corresponding to each work order type based on the correction script, correction field and constraint conditions corresponding to each work order type; Receive a process correction work order input by a user through the work order portrait template, and determine correction features corresponding to the process correction work order, wherein the correction features include the length of the correction script corresponding to the process correction work order, the number of data tables affected by the process correction work order and the total number of data rows involved, and the number of data modified in the key data table corresponding to the process correction work order; Determine multiple risk factors based on the length of the correction script, the number of data tables affected by the process correction work order, the total number of data rows involved in the process correction work order, and the number of data modifications in the key data table corresponding to the process correction work order; obtain the risk characteristic values corresponding to each risk factor in each historical correction work order based on the risk factors; normalize the risk characteristic values to obtain standardized risk characteristic values, and construct a training data set based on the standardized risk characteristic values; input the standardized risk characteristic values in the training data set into the initial risk identification model to obtain the risk prediction results corresponding to each historical correction work order; sample and generate a perturbation sample set based on the risk characteristic value corresponding to any historical correction work order; train a local explanation proxy model based on the perturbation sample set; interpret the risk prediction results output by the initial risk identification model based on the local explanation proxy model, and determine the contribution weight of each risk factor to the prediction result of a single historical correction work order; iteratively optimize the initial risk identification model based on the contribution weight to obtain a trained risk identification model; Based on the correction features, using the risk identification model, performing risk identification on the process correction work order, and determining the risk level corresponding to the process correction work order; Determining, according to the risk level, an approval node corresponding to the process correction work order, and determining, based on the approval node, an optimal approval path corresponding to the process correction work order; An approval process corresponding to the process correction work order is generated according to the optimal approval path.

7. A computer-readable storage medium, characterized in that The computer-readable storage medium stores one or more programs, which, when executed by an electronic device including multiple application programs, enable the electronic device to execute the process management method according to any one of claims 1 to 3.

8. A computer program product, characterized in that The invention comprises a computer program, which implements the process management method according to any one of claims 1 to 3 when executed by a processor.

Images