Method, device and equipment for training and using process timeout prediction model, and medium

Abstract

The application discloses a process timeout prediction model training and using method, device, equipment and medium. The process timeout prediction model training method comprises the following steps: constructing a process tree model according to a process timeout influence index; determining a process sample training set with label information according to the process tree model; and training a pre-constructed process timeout prediction model according to the process sample training set. The embodiment of the application improves the process timeout prediction result accuracy.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus, device, and medium for training and using a process timeout prediction model. Background Technology

[0002] Currently, indicator monitoring systems can only monitor or statistically analyze existing risk indicators, making it difficult to effectively predict future risk indicators. For example, providing early warnings for process timeouts is crucial in automated office scenarios. However, current self-learning capabilities for predicting process timeouts are poor, resulting in low prediction accuracy and limited effectiveness in real-world business applications. Summary of the Invention

[0003] This invention provides a method, apparatus, device, and medium for training and using a process timeout prediction model to improve the accuracy of process timeout prediction results.

[0004] According to one aspect of the present invention, a method for training a process timeout prediction model is provided, comprising:

[0005] Construct a process tree model based on the impact indicators of process timeout;

[0006] Based on the process tree model, determine the process sample training set with labeled information;

[0007] The pre-built process timeout prediction model is trained based on the process sample training set.

[0008] According to another aspect of the present invention, a process timeout prediction method is provided, comprising:

[0009] Retrieve the current process data for the newly initiated process;

[0010] The current process data is input into the trained process timeout prediction model to obtain the process timeout prediction result; wherein, the process timeout prediction model is trained using the training method of the process timeout prediction model provided in the embodiment of the present invention;

[0011] If the predicted timeout result is a timeout, then a timeout score is determined, and a corresponding timeout reminder is generated based on the timeout score.

[0012] According to another aspect of the present invention, a training apparatus for a process timeout prediction model is provided, the apparatus comprising:

[0013] The process tree model building module is used to build a process tree model based on the process timeout impact indicators;

[0014] The sample training set determination module is used to determine a process sample training set with label information based on the process tree model.

[0015] The model training module is used to train the pre-built process timeout prediction model based on the process sample training set.

[0016] According to another aspect of the present invention, a process timeout prediction apparatus is provided, the apparatus comprising:

[0017] The current process data acquisition module is used to acquire the current process data of a newly initiated process;

[0018] The prediction result determination module is used to input the current process data into the trained process timeout prediction model to obtain the process timeout prediction result; wherein, the process timeout prediction model is trained using the training method of the process timeout prediction model provided in the embodiment of the present invention;

[0019] The timeout score determination module is used to determine the process timeout score if the process timeout prediction result is a process timeout, and generate a corresponding timeout reminder based on the process timeout score.

[0020] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0021] At least one processor; and

[0022] A memory communicatively connected to the at least one processor; wherein,

[0023] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the training method and / or training of the process timeout prediction model according to any embodiment of the present invention.

[0024] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions, the computer instructions being configured to cause a processor to execute and implement the training method and / or training of the process timeout prediction model according to any embodiment of the present invention.

[0025] The present invention proposes a solution that constructs a process tree model based on process timeout impact indicators; determines a training set of process samples with tagged information based on the process tree model; and trains a pre-constructed process timeout prediction model based on the process sample training set. This achieves accurate training of the process timeout prediction model, enables automated prediction of process timeouts, and improves the accuracy of timeout prediction and alerts through the process timeout prediction model.

[0026] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a flowchart of a training method for a process timeout prediction model provided in Embodiment 1 of the present invention;

[0029] Figure 2 This is a flowchart of a training method for a process timeout prediction model provided in Embodiment 2 of the present invention;

[0030] Figure 3 This is a flowchart of a process timeout prediction method provided in Embodiment 3 of the present invention;

[0031] Figure 4 This is a schematic diagram of the structure of a training device for a process timeout prediction model according to Embodiment 4 of the present invention;

[0032] Figure 5 This is a schematic diagram of a process timeout prediction device provided in Embodiment 5 of the present invention;

[0033] Figure 6 This is a schematic diagram of the structure of an electronic device that implements the training method and / or process timeout prediction method of the process timeout prediction model in the embodiments of the present invention. Detailed Implementation

[0034] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0035] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0036] Example 1

[0037] Figure 1 This is a flowchart illustrating a training method for a process timeout prediction model according to Embodiment 1 of the present invention. This embodiment is applicable to predicting whether a business process will time out in an automated office setting. The method can be executed by a training device for the process timeout prediction model, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method includes:

[0038] S110. Construct a process tree model based on the process timeout impact indicators.

[0039] The impact of process timeout metrics varies depending on the specific application scenario. For example, in an OA (Office Automation) scenario, process timeout metrics can include personnel, processes, and departments. The personnel attribute level can include approvers and initiators; the process attribute level can include category, initiation time, and whether it's a form-based process; and the department attribute level can include department level and departmental section.

[0040] It should be noted that, regarding the impact indicators of process timeouts in OA scenarios, there are key impact indicators and redundant impact indicators. Specifically, decision tree methods and the k-means clustering algorithm can be used to analyze the key impact indicators of process timeouts in OA scenarios.

[0041] For example, relevant technical personnel can pre-construct the influencing indicators and weights of OA process timeouts using a tree structure. The weights can be pre-set by the technical personnel according to actual needs. Using the decision tree method, based on the constructed tree structure, entropy values ​​are calculated for the influencing indicators of process timeouts related to personnel, processes, and departments. Based on the calculation results, pruning operations are performed on the decision tree edges to obtain the optimal tree structure, which is then used as the process tree model. The model structure of the process tree model can be shown in Table 1.

[0042] Table 1

[0043]

[0044] S120. Based on the process tree model, determine the training set of process samples with label information.

[0045] For example, process data within a preset time period can be used as a process sample training set based on the process timeout impact indicators in the process tree model and the attribute levels corresponding to each timeout impact indicator. The preset time period can be pre-set by relevant technical personnel, for example, three months. The label information can include process timeout and process not timed out.

[0046] Specifically, based on the process data model, process data within a preset time period can be obtained. Then, using a preset K-means clustering algorithm, the process data is classified according to preset process categories, resulting in at least one set of process categories with different characteristics. Each set of process categories may include at least one OA (Office Automation) process, and each OA process corresponds to a timeout or non-timeout label. The set of process categories with these labels is used as the process sample training set. The label information in the process category set can be predetermined by relevant technical personnel.

[0047] S130. Train the pre-built process timeout prediction model based on the process sample training set.

[0048] For example, the process sample training set can be input into a pre-built process timeout prediction model to obtain the timeout prediction result of the model; the process timeout prediction model can be trained based on the timeout prediction result and the actual result in the label information of the process sample training set until the training deadline is met, and then the trained process timeout prediction model is obtained.

[0049] The process timeout prediction model can be pre-set by relevant technical personnel. For example, the process timeout prediction model can be an SVM (Support Vector Machine). The training cutoff condition can be that the loss value determined by the loss function no longer changes, or the change range is stable, based on the predicted and actual results, which is considered to meet the training cutoff condition.

[0050] The present invention proposes a solution that constructs a process tree model based on process timeout impact indicators; determines a training set of process samples with tagged information based on the process tree model; and trains a pre-constructed process timeout prediction model based on the process sample training set. This achieves accurate training of the process timeout prediction model, enables automated prediction of process timeouts, and improves the accuracy of timeout prediction and alerts through the process timeout prediction model.

[0051] Example 2

[0052] Figure 2 This is a flowchart of a training method for a process timeout prediction model provided in Embodiment 2 of the present invention. This embodiment is an optimization and improvement based on the above technical solutions.

[0053] Furthermore, the step "constructing a process tree model based on process timeout impact indicators" is refined into "determining at least one process timeout impact indicator, and at least one impact attribute indicator corresponding to each process timeout impact indicator; constructing at least one process branch structure corresponding to each process timeout impact indicator based on each impact attribute indicator; and constructing a process tree model based on each process branch structure corresponding to each process timeout impact indicator." This improves the method for constructing the process tree model.

[0054] like Figure 2 As shown, the method includes the following specific steps:

[0055] S210. Determine at least one process timeout impact indicator, and at least one impact attribute indicator corresponding to each process timeout impact indicator.

[0056] The impact indicators for process timeouts can be determined based on the business scenario. In an OA application scenario, these indicators can include personnel, processes, and departments. The impact attribute indicators corresponding to these indicators can represent the attribute information of the indicator. For example, the impact attribute indicators for personnel can include approvers and initiators. Approvers and initiators can be considered second-level branches of personnel. The impact attribute indicators for approvers can include gender, age, position, and department, which can be considered third-level branches. The impact attribute indicators for departments can include department level and department section, which can be considered fourth-level branches. This embodiment will not elaborate further on these aspects.

[0057] It's important to note that in OA (Office Automation) applications, there are multiple metrics affecting process timeouts, including both critical and redundant metrics. Critical metrics are those that definitely impact process timeouts, such as the age of the approver. Redundant metrics, on the other hand, do not affect process timeouts, such as the name of the approver. Therefore, redundant metrics can be manually removed by technical personnel or automatically.

[0058] S220. Based on the impact attribute indicators corresponding to the timeout impact indicators of each process, construct at least one process branch structure corresponding to the timeout impact indicators of each process.

[0059] Understandably, in decision tree methods, entropy can represent a measure of the uncertainty of a random variable, that is, the degree of disorder within an object, and can be used to determine the quality of selecting a data point as a tree node. The smaller the entropy, the less disorder; the less disorder, the better the tree node. Therefore, it is possible to construct a process branch structure corresponding to the timeout impact indicators of each process.

[0060] For example, let's illustrate the impact of process timeouts on personnel. Process branch structure A: Personnel → Gender, Personnel → Age; Process branch structure B: Personnel → Salary, Personnel → Gender, Personnel → Age. The entropy of the constructed process branch structure is determined, and the next level is constructed based on the entropy value.

[0061] S230. Construct a process tree model based on the process branch structure corresponding to the timeout impact indicators of each process.

[0062] For example, the entropy value of each process branch structure can be determined based on the timeout impact indicators of each process, and then a process tree model can be constructed based on the entropy value of each process branch structure.

[0063] In one optional embodiment, a process tree model is constructed based on the process branch structure corresponding to each process timeout impact indicator, including: determining the structural entropy value corresponding to each process branch structure corresponding to each process timeout impact indicator; determining the target process structure corresponding to the corresponding process timeout impact indicator based on the structural entropy value; and constructing a process tree model based on the target process structure corresponding to each process timeout indicator.

[0064] For example, if the first-level process branch structure for personnel is A: Personnel → Position, Personnel → Age; and the process branch structure B: Personnel → Salary, Personnel → Position, Personnel → Age, the entropy value of the determined process branch structure A is 1, and the entropy value of the process branch structure B is 100. Therefore, the "Salary" branch can be removed. After removal, the second-level process branch structure for personnel is constructed: process branch structure Aa: Personnel → Position → Technical Position, Personnel → Position → Audit Position; Ab: Personnel → Position → Technical Position, Personnel → Position → Audit Position, Personnel → Position → Administrative Position. The entropy value of Aa is 1, and the entropy value of Ab is 100. Therefore, the "Administrative Position" branch can be removed. Similarly, for the subsequently created process branch structures, the target process structure can be determined step by step from top to bottom. The target process structures corresponding to each process timeout index are combined to obtain a process tree model. Among them, the constructed process tree model is the model with the lowest entropy value.

[0065] For example, the structural entropy value corresponding to the process branch structure can be determined as follows:

[0066]

[0067] Where S is the structural entropy value corresponding to the process branch structure; n indicates that there are n groups of process data with determined entropy values; m indicates that m groups of process data time out of the n groups.

[0068] For example, by comparing the structural entropy values ​​of each process branch structure such as personnel, processes and departments in the established tree structure, and combining the operation of pruning while building the decision tree, the optimal tree structure can be obtained as shown in Table 2.

[0069] Table 2

[0070]

[0071] S240. Based on the process tree model, determine the training set of process samples with label information.

[0072] In an optional embodiment, determining a process sample training set with label information according to the process tree model includes: acquiring process data related to each influence attribute indicator within a preset time period based on the influence attribute indicators in the process tree model, and constructing a process dataset; dividing the process dataset into at least one target data set based on the attribute information of the process data in the process dataset; wherein the attribute information of the process data is related to the influence attribute indicators; determining the label category corresponding to each target data set, and determining each target data set with label category as the process sample training set.

[0073] The preset time period can be three months. For example, process data corresponding to various influencing attribute indicators can be obtained. Examples include process data such as the name, age, job title, department / section level, and department / office of the approver. A process dataset is constructed by acquiring process data from the past three months.

[0074] The process dataset can include at least one process and its corresponding timeout result. For example, the process could be an approval process where an approver from a certain name, gender, and department submits a leave application at a certain time, and the corresponding approval result is "process approval timed out." The target dataset can include at least one approval process and its corresponding approval result, with the approval result serving as the label category for that approval process. The label categories include "process timed out" and "process did not time out." The target dataset with these label categories is then used as the process sample training set.

[0075] In an optional embodiment, the process dataset is divided into at least one target data set based on the attribute information of the process data in the process dataset, including: selecting at least one target attribute indicator from each influencing attribute indicator, and using the process data corresponding to each target attribute indicator as the target centroid data; determining the Euclidean distance between other process data in the process dataset and each target centroid data; wherein, the other process data are process data other than the target centroid data; and determining at least one target data set based on each Euclidean distance.

[0076] The target attribute indicator can be a process category. Correspondingly, the process data corresponding to each process category can be used as the target centroid data. The process data is then classified using the K-means clustering algorithm.

[0077] For example, for each point in other process data, calculate its Euclidean distance to each target centroid data point; that is, determine which target centroid it is closer to, and then assign it to the data set belonging to that target centroid. After all process data are grouped into data sets, recalculate the centroid of each data set. If the distance between the newly calculated centroid and the original centroid is less than a set fixed threshold, the process is terminated, and the classified process data is used as the target data set. If the distance between the new centroid and the original centroid changes significantly, the set partitioning step is repeated. After obtaining the process classification results, each process in the target data set is labeled with a result tag to obtain a label category.

[0078] S250. Train the pre-built process timeout prediction model based on the process sample training set.

[0079] In an optional embodiment, before training the pre-built process timeout prediction model based on the process sample training set, the method further includes: determining model initialization parameters based on the current business scenario; wherein the model initialization parameters include at least one of a kernel function, a penalty parameter, a loss function, and a model iteration termination condition threshold. Correspondingly, training the pre-built process timeout prediction model based on the process sample training set includes: training the pre-built process timeout prediction model based on the process sample training set and the model initialization parameters.

[0080] The model initialization parameters may include kernel functions, penalty parameters, loss functions, and model iteration termination condition thresholds, etc.

[0081] Among these, the kernel function can preferably be a robust radial basis function (RBF) kernel, i.e., a Gaussian kernel function. Robust RBF kernels have good resistance to noise in the data, and their parameters determine the function's effective range. The kernel function can be calculated as follows:

[0082]

[0083] Where, x i This represents the input value of the data variable after digitizing the indicators affecting process timeouts (such as staff age and process classification). i σ represents the radial basis function center vector of the hidden nodes in the process timeout prediction model; σ represents the connection weights between the hidden layers and output nodes of the process timeout prediction network model. σ can be calculated experimentally using test data, such as exhaustive testing, to optimize its performance without overfitting.

[0084] This invention's embodiments achieve accurate construction of the process tree model by determining at least one impact attribute indicator corresponding to each process timeout impact indicator; constructing at least one process branch structure corresponding to each process timeout impact indicator based on the impact attribute indicators; and building a process tree model based on the process branch structures corresponding to each process timeout impact indicator. This improves the accuracy of determining the process sample training set. Furthermore, by constructing a process dataset and dividing it into at least one target data set based on the attribute information of the process data, and determining the label category corresponding to each target data set, the method for determining the process sample training set is further improved, thereby further improving the training accuracy of the model and ultimately improving the prediction accuracy of process timeouts.

[0085] Example 3

[0086] Figure 3 This is a flowchart of a process timeout prediction method provided in Embodiment 3 of the present invention. This embodiment is applicable to predicting whether a business process in an automated office scenario will time out. The method can be executed by a process timeout prediction device, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 3 As shown, the method includes:

[0087] S310. Obtain the current process data for the newly initiated process.

[0088] The current process data can be the process data corresponding to the timeout impact indicators of each newly initiated process.

[0089] S320. Input the current process data into the trained process timeout prediction model to obtain the process timeout prediction result.

[0090] For example, the current process data is input into a trained process timeout prediction model to obtain a process timeout prediction result, wherein the process timeout prediction result is either the process has timed out or the process has not timed out. The process timeout prediction model is trained using the training method of the process timeout prediction model in this embodiment of the invention.

[0091] S330. If the predicted timeout result is a timeout, determine the timeout score and generate a corresponding timeout reminder based on the timeout score.

[0092] It should be noted that, in determining the process timeout prediction result based on the current process data, the process timeout prediction model, being a binary classification problem, obtains corresponding classification probability values. The model then outputs the classification result by comparing these probability values ​​with a preset threshold. If the classification probability value is greater than the preset threshold, the process is considered to have timed out; if the classification probability value is not greater than the preset threshold, the process is considered not to have timed out.

[0093] The process timeout score can be determined using a classification probability value and a preset threshold. When the classification result is a process timeout, the ratio between the preset threshold and the classification probability value can be used as the process timeout score; alternatively, the percentage difference between the classification probability value and the preset threshold can also be used as the process timeout score. A corresponding timeout reminder is then generated based on the process timeout score.

[0094] Optionally, a corresponding level can be generated based on the average score fluctuation of different process timeout scores. For example, Level I is an average score fluctuation within 30%; Level II is an average score fluctuation within 50%; Level III is an average score fluctuation within 70%; and Level IV is an average score fluctuation of more than 70%.

[0095] The present invention obtains the current process data of a newly initiated process; inputs the current process data into a trained process timeout prediction model to obtain a process timeout prediction result; if the process timeout prediction result is a process timeout, a process timeout score is determined, and a corresponding timeout reminder is generated based on the process timeout score. This improves the accuracy of process timeout prediction and enables automated prediction of subsequent process timeouts. By using a process timeout prediction model to predict and remind users of process timeouts, the accuracy of the prediction results is improved.

[0096] Example 4

[0097] Figure 4 This is a schematic diagram of the structure of a training device for a process timeout prediction model provided in Embodiment 4 of the present invention. The training device for a process timeout prediction model provided in this embodiment of the present invention is applicable to predicting whether a business process will time out in an automated office scenario. This training device for the process timeout prediction model can be implemented in hardware and / or software, such as... Figure 4 As shown, the device specifically includes: a process tree model construction module 401, a sample training set determination module 402, and a model training module 403. Among them,

[0098] The process tree model construction module 401 is used to construct a process tree model based on the process timeout impact indicators.

[0099] The sample training set determination module 402 is used to determine a process sample training set with label information based on the process tree model.

[0100] The model training module 403 is used to train the pre-built process timeout prediction model based on the process sample training set.

[0101] The present invention proposes a solution that constructs a process tree model based on process timeout impact indicators; determines a training set of process samples with tagged information based on the process tree model; and trains a pre-constructed process timeout prediction model based on the process sample training set. This achieves accurate training of the process timeout prediction model, enables automated prediction of process timeouts, and improves the accuracy of timeout prediction and alerts through the process timeout prediction model.

[0102] Optionally, the process tree model construction module 401 includes:

[0103] An impact indicator determination unit is used to determine at least one process timeout impact indicator, and at least one impact attribute indicator corresponding to each process timeout impact indicator.

[0104] A branch structure construction unit is used to construct at least one process branch structure corresponding to each process timeout impact indicator based on each impact attribute indicator corresponding to each process timeout impact indicator.

[0105] The process tree model construction unit is used to construct the process tree model based on the process branch structure corresponding to each process timeout impact indicator.

[0106] Optionally, the process tree model construction unit includes:

[0107] The structural entropy value determination subunit is used to determine the structural entropy value of each process branch structure corresponding to each process timeout impact index.

[0108] The target process structure determination sub-unit is used to determine the target process structure corresponding to the timeout impact index of the corresponding process based on the entropy value of each structure.

[0109] The process tree model construction subunit is used to construct a process tree model based on the target process structure corresponding to each process timeout indicator.

[0110] Optionally, the sample training set determination module 402 includes:

[0111] The process dataset construction unit is used to obtain process data related to each of the impact attribute indicators within a preset time period based on the impact attribute indicators in the process tree model, and to construct the process dataset.

[0112] The target data set partitioning unit is used to partition the process dataset into at least one target data set based on the attribute information of the process data in the process dataset; wherein the attribute information of the process data is related to the influencing attribute index;

[0113] The sample training set determination unit is used to determine the label category corresponding to each of the target data sets, and to determine each target data set with the label category as the process sample training set.

[0114] Optionally, the target data set partitioning unit includes:

[0115] The target centroid data determination subunit is used to select at least one target attribute indicator from each of the said influencing attribute indicators, and to use the process data corresponding to each of the said target attribute indicators as the target centroid data.

[0116] The Euclidean distance determination subunit is used to determine the Euclidean distance between other process data in the process dataset and each of the target centroid data; wherein, the other process data are process data other than the target centroid data;

[0117] The target data set determination subunit is used to determine at least one target data set based on each of the Euclidean distances.

[0118] The training device for the process timeout prediction model provided in this embodiment of the invention can execute the training method for the process timeout prediction model provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0119] Example 5

[0120] Figure 5 This is a schematic diagram of a process timeout prediction device provided in Embodiment 5 of the present invention. The process timeout prediction device provided in this embodiment of the present invention is applicable to predicting whether a business process in an automated office scenario will time out. This process timeout prediction device can be implemented in hardware and / or software, such as… Figure 5 As shown, the device specifically includes: a current process data acquisition module 501, a prediction result determination module 402, and a timeout score determination module 403. Among them,

[0121] The current process data acquisition module 501 is used to acquire the current process data of a newly initiated process;

[0122] The prediction result determination module 502 is used to input the current process data into the trained process timeout prediction model to obtain the process timeout prediction result; wherein, the process timeout prediction model is trained using the training method of the process timeout prediction model described in the embodiment of the present invention;

[0123] The timeout score determination module 503 is used to determine the process timeout score if the process timeout prediction result is a process timeout, and generate a corresponding timeout reminder based on the process timeout score.

[0124] The present invention obtains the current process data of a newly initiated process; inputs the current process data into a trained process timeout prediction model to obtain a process timeout prediction result; if the process timeout prediction result is a process timeout, a process timeout score is determined, and a corresponding timeout reminder is generated based on the process timeout score. This improves the accuracy of process timeout prediction and enables automated prediction of subsequent process timeouts. By using a process timeout prediction model to predict and remind users of process timeouts, the accuracy of the prediction results is improved.

[0125] The process timeout prediction device provided in this embodiment of the invention can execute the process timeout prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0126] Example 6

[0127] Figure 6 A schematic diagram of an electronic device 60 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0128] like Figure 6 As shown, the electronic device 60 includes at least one processor 61 and a memory, such as a read-only memory (ROM) 62 and a random access memory (RAM) 63, communicatively connected to the at least one processor 61. The memory stores computer programs executable by the at least one processor. The processor 61 can perform various appropriate actions and processes based on the computer program stored in the ROM 62 or loaded into the RAM 63 from storage unit 68. The RAM 63 may also store various programs and data required for the operation of the electronic device 60. The processor 61, ROM 62, and RAM 63 are interconnected via a bus 64. An input / output (I / O) interface 65 is also connected to the bus 64.

[0129] Multiple components in electronic device 60 are connected to I / O interface 65, including: input unit 66, such as keyboard, mouse, etc.; output unit 67, such as various types of monitors, speakers, etc.; storage unit 68, such as disk, optical disk, etc.; and communication unit 69, such as network card, modem, wireless transceiver, etc. Communication unit 69 allows electronic device 60 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0130] Processor 61 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 61 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 61 performs the various methods and processes described above, such as training methods for process timeout prediction models and / or process timeout prediction methods.

[0131] In some embodiments, the training method and / or process timeout prediction method of the process timeout prediction model may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 68. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 60 via ROM 62 and / or communication unit 69. When the computer program is loaded into RAM 63 and executed by processor 61, one or more steps of the training method and / or process timeout prediction method of the process timeout prediction model described above may be performed. Alternatively, in other embodiments, processor 61 may be configured to execute the training method and / or process timeout prediction method of the process timeout prediction model by any other suitable means (e.g., by means of firmware).

[0132] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0133] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0134] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0135] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0136] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0137] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0138] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0139] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A training method for a process timeout prediction model, characterized in that, include: A process tree model is constructed based on the process timeout impact indicators; wherein, the process timeout impact indicators include critical process timeout impact indicators and redundant process timeout impact indicators; the critical process timeout impact indicators refer to those that can definitely have an impact on process timeout; the redundant process timeout impact indicators refer to those that do not have an impact on process timeout. Based on the process tree model, determine the process sample training set with labeled information; The pre-built process timeout prediction model is trained based on the process sample training set. The step of constructing a process tree model based on process timeout impact indicators includes: Identify at least one process timeout impact indicator, and at least one impact attribute indicator corresponding to each process timeout impact indicator; Based on the impact attribute indicators corresponding to each of the process timeout impact indicators, construct at least one process branch structure corresponding to each of the process timeout impact indicators. Based on the process branch structure corresponding to each of the process timeout impact indicators, construct the process tree model; The step of constructing the process tree model based on the process branch structure corresponding to each of the process timeout impact indicators includes: Determine the structural entropy value of each process branch structure corresponding to each process timeout impact index; wherein, the structural entropy value is used to judge the quality of a data as a tree node, and the smaller the structural entropy value, the better the tree node; Based on the entropy values ​​of each structure, determine the target process structure corresponding to the timeout impact index of the corresponding process. Based on the target process structure corresponding to each of the process timeout impact indicators, a process tree model is constructed; wherein, the constructed process tree model is the model with the lowest entropy value.

2. The method according to claim 1, characterized in that, The step of determining the training set of process samples with label information based on the process tree model includes: Based on the impact attribute indicators in the process tree model, process data related to each impact attribute indicator within a preset time period is obtained to construct a process dataset. Based on the attribute information of the process data in the process dataset, the process dataset is divided into at least one target data set; wherein, the attribute information of the process data is related to the influencing attribute index; Determine the label category corresponding to each of the target data sets, and determine each target data set with the label category as the process sample training set.

3. The method according to claim 2, characterized in that, The step of dividing the process dataset into at least one target data set based on the attribute information of the process data in the process dataset includes: Select at least one target attribute indicator from the aforementioned influencing attribute indicators, and use the process data corresponding to each target attribute indicator as the target centroid data; Determine the Euclidean distances between other process data in the process dataset and each of the target centroid data; wherein, the other process data refers to process data other than the target centroid data; Based on the Euclidean distances described above, at least one target dataset is determined.

4. A method for predicting process timeouts, characterized in that, include: Retrieve the current process data for the newly initiated process; The current process data is input into the trained process timeout prediction model to obtain the process timeout prediction result; wherein the process timeout prediction model is trained using the training method of the process timeout prediction model according to any one of claims 1-3; If the predicted timeout result is a timeout, then a timeout score is determined, and a corresponding timeout reminder is generated based on the timeout score.

5. A training device for a process timeout prediction model, characterized in that, include: The process tree model construction module is used to construct a process tree model based on process timeout impact indicators. These indicators include key process timeout impact indicators and redundant process timeout impact indicators. Key process timeout impact indicators refer to those that definitely have an impact on process timeout; redundant process timeout impact indicators refer to those that do not have an impact on process timeout. The sample training set determination module is used to determine a process sample training set with label information based on the process tree model. The model training module is used to train the pre-built process timeout prediction model based on the process sample training set. The process tree model construction module includes: An impact indicator determination unit is used to determine at least one process timeout impact indicator, and at least one impact attribute indicator corresponding to each process timeout impact indicator. A branch structure construction unit is used to construct at least one process branch structure corresponding to each process timeout impact indicator based on each impact attribute indicator corresponding to each process timeout impact indicator. The process tree model construction unit is used to construct the process tree model based on the process branch structure corresponding to each of the process timeout impact indicators. The process tree model construction unit includes: The structure entropy value determination subunit is used to determine the structure entropy value of each process branch structure corresponding to each process timeout impact index; wherein, the structure entropy value is used to judge the quality of a data as a tree node, and the smaller the structure entropy value, the better the tree node; The target process structure determination sub-unit is used to determine the target process structure corresponding to the timeout impact index of the corresponding process based on the entropy value of each structure. The process tree model construction subunit is used to construct a process tree model based on the target process structure corresponding to each process timeout impact index; wherein, the constructed process tree model is the model with the lowest entropy value.

6. A process timeout prediction device, characterized in that, include: The current process data acquisition module is used to acquire the current process data of a newly initiated process; The prediction result determination module is used to input the current process data into the trained process timeout prediction model to obtain the process timeout prediction result; wherein, the process timeout prediction model is trained using the training method of the process timeout prediction model according to any one of claims 1-3; The timeout score determination module is used to determine the process timeout score if the process timeout prediction result is a process timeout, and generate a corresponding timeout reminder based on the process timeout score.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the training method of the process timeout prediction model according to any one of claims 1-3, and / or perform the process timeout prediction method according to claim 4.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the training method of the process timeout prediction model according to any one of claims 1-3, and / or execute the process timeout prediction method according to claim 4.

Images