Flow link evaluation method based on graph neural network

Abstract

The invention provides a flow link evaluation method based on a graph neural network, and the method comprises the steps: constructing a planning flow chart of the work according to work data uploaded by a user, correcting a graph operator corresponding to each flow node through the graph neural network, and generating a plurality of logic cycle conditions, performing logic analysis on the working data to obtain node results corresponding to the flow nodes, capturing flow information contained in each node result by using a graph neural network to obtain corresponding graph structure information, analyzing investment per cost, and establishing overall graph structure information of the planning flow chart; bottleneck information and redundant information contained in the overall graph structure information are utilized to determine bottleneck process nodes and redundant process nodes, a cost calibration suggestion is generated, each bottleneck process node and each redundant process node are respectively positioned in a planning flow chart, a node modification suggestion is generated, and a cost calibration suggestion is generated; the method is used for helping to evaluate the rationality of the process link design in the development and design stage.

Description

Technical Field

[0001] The present invention relates to the technical field of business process management and processing, and in particular to a process link evaluation method based on graph neural network. Background Art

[0002] With the advent of the global digital era, corporate businesses are facing new and greater challenges. In order to cope with global competitive pressure and improve the ability of enterprises to cope with various uncertainties, enterprises often need to dynamically adjust their development strategies and business scope according to market changes and needs, and then change, reconstruct and optimize their business processes to improve their competitiveness. In the face of various changes and uncertainties, how to ensure that business processes can meet needs has become an important condition for corporate development and gaining competitive advantages. The rationality assessment of business processes is one of the important evaluation indicators. Process rationality refers to the ability of the business to ensure the completion of tasks when responding to threats and uncertainties. The rationality of the process directly reflects the level of success of the business process in the face of uncertain environments.

[0003] At present, for such situations, we can only analyze the rationality of the flowchart based on the indicator data of each link after the flowchart has been running for a period of time. Obviously, its efficiency is low, the update speed is slow, and it may even affect the stable operation of the enterprise business.

[0004] Therefore, the present invention provides a process link evaluation method based on graph neural network. Summary of the invention

[0005] The present invention provides a process link evaluation method based on graph neural network, which is used to help evaluate the rationality of process link design in the development and design stage.

[0006] The present invention provides a process link evaluation method based on graph neural network, comprising:

[0007] Step 1: Build a planning flowchart for this work based on the work data uploaded by the user, and use the graph neural network to correct the graph operator corresponding to each process node in the planning flowchart;

[0008] Step 2: Generate several logical loop conditions of the planning flow chart according to the graph operator, use the logical loop conditions to perform logical analysis on the work data, and obtain node results corresponding to process nodes;

[0009] Step 3: Use the graph neural network to capture the process information contained in each of the node results, obtain the graph structure information corresponding to each of the process nodes, and analyze the cost input corresponding to each of the process nodes;

[0010] Step 4: Establish the overall graph structure information of the planning flowchart, use the graph neural network to screen the bottleneck information and redundant information contained in the overall graph structure information, and determine the bottleneck process nodes and redundant process nodes of the planning flowchart;

[0011] Step 5: Generate cost calibration suggestions based on the cost input corresponding to each of the process nodes, locate each of the bottleneck process nodes and each of the redundant process nodes in the planning flowchart, and generate node modification suggestions.

[0012] In one practicable manner,

[0013] The step 1 comprises:

[0014] Step 11: Obtain the work data uploaded by the user, classify the work data to obtain the process data segment and execution data segment corresponding to this work, establish a matching relationship between the process data segment and the execution data segment, and generate several process links of this work;

[0015] Step 12: Analyze the interaction relationship between different process data segments, generate corresponding interaction conditions, match the corresponding process links for each interaction condition, use the interaction conditions to deploy data for the corresponding execution data segments, and use the data range corresponding to each process link to adjust the corresponding process link and generate a planning flowchart;

[0016] Step 13: using the graph neural network to identify the point representation vector corresponding to each process node in the planning flowchart, and respectively obtaining the node attributes corresponding to each process node, determining the interaction logic corresponding to the point representation vector according to the node attributes, and generating a plurality of interaction representation vectors of the planning flowchart;

[0017] Step 14: Determine the basic operator information of the corresponding process node according to the point representation vector, determine the deep operator information of the corresponding process node according to the interaction representation vector, match the corresponding operator elements for the process node according to the basic operator information, match the corresponding operator conditions for the operator elements according to the deep operator information, and generate a graph operator corresponding to each of the process nodes.

[0018] In one practicable manner,

[0019] The step 2 comprises:

[0020] Step 21: generating a plurality of process directions of the planning flowchart according to the graph operator, dividing each of the process directions into a plurality of unit stages, respectively counting a plurality of unit stages corresponding to each process node, and generating a logic loop condition corresponding to each process node in the planning flowchart based on the process direction corresponding to each unit stage;

[0021] Step 22: Establish a logic model of the planning flowchart using the logic loop condition, input the work data into the logic model for logic analysis, generate a process result corresponding to each process trend, and obtain the data change process of the work data in the logic model;

[0022] Step 23: Based on the data change process and the process results corresponding to each process direction, determine the data processing characteristics and data processing results corresponding to each unit stage in the planning flowchart, generate node results of corresponding process nodes and display them.

[0023] In one practicable manner,

[0024] Also includes:

[0025] Obtaining the work target characteristics uploaded by the user;

[0026] Filter the output node results corresponding to the output process nodes and generate process result features;

[0027] When the process result characteristics are inconsistent with the work target characteristics, the planning flowchart is determined to be abnormal, and an abnormal reminder is generated and displayed.

[0028] In one practicable manner,

[0029] The step 3 comprises:

[0030] Step 31: input each of the node results into the graph neural network respectively, identify the local structure between different node results, generate an adjacency list of the planning flowchart in combination with the presentation data corresponding to each of the node results, and capture the process information corresponding to each of the process nodes in the adjacency list;

[0031] Step 32: Establishing the global structure of the planning flowchart according to the local structure between different node results, inputting the process information into the global structure, obtaining the planning role of each process node in the planning flowchart, and establishing the graph structure information of the corresponding process node according to the planning role;

[0032] Step 33: Determine the executable workload of the corresponding process node and the transferable workload between different process nodes according to the graph structure information, and determine the cost investment corresponding to each of the process nodes.

[0033] In one practicable manner,

[0034] The step 4 comprises:

[0035] Step 41: Map the graph structure information corresponding to each of the process nodes in the planning flowchart, add a corresponding external structure to each of the graph structure information according to the mapping result, and connect the graph structure information using the external structure to generate the overall graph structure information of the planning flowchart;

[0036] Step 42: traverse the overall graph structure information using the graph neural network to obtain a number of process decision statements of the planning flowchart, generate a decision tree of the planning flowchart, analyze the decision execution information corresponding to each leaf node in the decision tree, and screen and determine the overloaded leaf nodes with bottleneck information and the isolated leaf nodes with redundant information;

[0037] Step 44: Obtain the overload graph structure information corresponding to the overload leaf node, construct the execution process corresponding to the overload leaf node according to the overload graph structure information, select the corresponding experimental data from the work data and input it into the execution process to check whether the execution time of the corresponding overload leaf node is abnormal, and determine that the process node corresponding to the overload leaf node is a bottleneck process node;

[0038] Step 45: Obtain the isolated graph structure information corresponding to the isolated node, identify the input execution object and the output execution object corresponding to the isolated leaf node in the isolated graph structure information, and when the isolated leaf node does not contain an output execution object or does not contain an output execution object, determine that the process node corresponding to the isolated node is a redundant process node.

[0039] In one practicable manner,

[0040] The process of screening and determining overloaded leaf nodes with bottleneck information and isolated leaf nodes with redundant information includes:

[0041] Obtaining input-output relationships between different leaf nodes in the decision tree, and establishing a plurality of leaf node flows;

[0042] Respectively obtain the last leaf node corresponding to each of the leaf node streams, and add an isolated identifier to each of the last leaf nodes;

[0043] The node input-output duration corresponding to each leaf node flow and the mean square error duration corresponding to each leaf node flow are obtained respectively, the overloaded leaf nodes whose node input-output duration is higher than the corresponding mean square error duration are obtained, and a bottleneck identifier is added to each overloaded leaf node.

[0044] In one practicable manner,

[0045] The step 5 comprises:

[0046] Step 51: respectively obtain the working pressure characteristics corresponding to each of the process nodes, use the working pressure characteristics to evaluate the rationality of the cost input corresponding to each of the process nodes, determine the first process node whose cost input is higher than the corresponding use cost, and the second process node whose cost input is lower than the corresponding use cost, and generate cost calibration suggestions;

[0047] Step 52: Locate each of the bottleneck process nodes and the redundant process nodes of the purpose in the planning flowchart, obtain the pressure to be diverted corresponding to each of the bottleneck process nodes, and generate node modification suggestions in combination with deletion suggestions for each of the redundant process nodes.

[0048] In one practicable manner,

[0049] Also includes:

[0050] A scoring report of the planning flowchart is generated and displayed according to the cost calibration suggestion and the node modification suggestion.

[0051] In one practicable manner,

[0052] Also includes:

[0053] The graph neural network is used to evaluate several node functions corresponding to each of the process nodes, and a function scoring report for each of the process nodes is generated and displayed.

[0054] The achievable beneficial effects of the above technical solution are as follows: in order to set a flowchart that meets the needs of users, improve the work efficiency and work quality of users, and reduce the errors of automatically generated flowcharts, first, a planning flowchart is built according to the work data uploaded by the user, and the graph neural network is used to correct the graph operator of each process node in the planning flowchart, and the planning flowchart is further divided into a number of logical loop conditions, and the node results of the process nodes are determined by logical analysis. The graph neural network is used again to capture the process information contained in each node result, and then the graph structure information of each process node is constructed. By constructing the graph structure information, the actual working structure of each process node can be visualized, and the working process of each process node can be intuitively understood, thereby determining the cost investment of each process node. Further, by constructing the overall graph structure information of the planning flowchart, the redundant information and bottleneck information contained in the planning flowchart are analyzed, and finally, the cost calibration suggestions and node modification suggestions of the planning flowchart are generated. In this way, each node in the planning flowchart can be analyzed, the unreasonable layout in the planning flowchart can be identified, and corresponding solutions can be proposed. Users can adjust the planning flowchart according to the given solution, which can not only quickly complete the adjustment purpose, but also adjust the structure of the planning flowchart appropriately according to the company's own conditions to achieve human-machine balance.

[0055] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0056] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0058] Figure 1 A schematic diagram of a workflow of a process link evaluation method based on a graph neural network in an embodiment of the present invention;

[0059] Figure 2 This is a workflow diagram of step 2 of a process link evaluation method based on a graph neural network in an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0061] Example 1

[0062] This embodiment provides a process link evaluation method based on graph neural network, such as Figure 1 As shown, including:

[0063] Step 1: Build a planning flowchart for this work based on the work data uploaded by the user, and use the graph neural network to correct the graph operator corresponding to each process node in the planning flowchart;

[0064] Step 2: Generate several logical loop conditions of the planning flow chart according to the graph operator, use the logical loop conditions to perform logical analysis on the work data, and obtain node results corresponding to process nodes;

[0065] Step 3: Use the graph neural network to capture the process information contained in each of the node results, obtain the graph structure information corresponding to each of the process nodes, and analyze the cost input corresponding to each of the process nodes;

[0066] Step 4: Establish the overall graph structure information of the planning flowchart, use the graph neural network to screen the bottleneck information and redundant information contained in the overall graph structure information, and determine the bottleneck process nodes and redundant process nodes of the planning flowchart;

[0067] Step 5: Generate cost calibration suggestions based on the cost input corresponding to each of the process nodes, locate each of the bottleneck process nodes and each of the redundant process nodes in the planning flowchart, and generate node modification suggestions.

[0068] In this example, the work data refers to the data uploaded by the user regarding the business operation of the enterprise;

[0069] In this example, the planning flowchart represents a flowchart generated by arranging the steps of this work in a process;

[0070] In this example, graph operators represent the conditions required for process nodes to perform different operations;

[0071] In this example, the logical loop condition represents a logical or cyclic execution condition in the planning flow chart;

[0072] In this example, the node result contains the information generated by the corresponding process node;

[0073] In this example, the graph structure information represents the structure of several process operations involved in a process node;

[0074] In this example, the overall graph structure information represents the structure of all process nodes involved in the planning flow chart;

[0075] In this example, the bottleneck information indicates the information generated when the process progress of the planning flowchart is slow;

[0076] In this example, redundant information refers to information in the planning flow chart that is redundant and has no specific meaning.

[0077] The working principle and beneficial effects of the above technical solution: In order to set a flowchart that meets the needs of users, improve the work efficiency and work quality of users, and reduce the errors of automatically generated flowcharts, first, a planning flowchart is built according to the work data uploaded by the user, and the graph neural network is used to correct the graph operator of each process node in the planning flowchart, and the planning flowchart is further divided into a number of logical loop conditions. The node results of the process nodes are determined by logical analysis, and the graph neural network is used again to capture the process information contained in each node result, and then the graph structure information of each process node is constructed. By constructing the graph structure information, the actual working structure of each process node can be visualized, and the working process of each process node can be intuitively understood, thereby determining the cost investment of each process node. Further, by building the overall graph structure information of the planning flowchart, the redundant information and bottleneck information contained in the planning flowchart are analyzed, and finally the cost calibration suggestions and node modification suggestions of the planning flowchart are generated. In this way, each node in the planning flowchart can be analyzed, the unreasonable layout in the planning flowchart can be identified, and corresponding solutions can be proposed. Users can adjust the planning flowchart according to the given solutions, which can not only quickly complete the adjustment purpose, but also adjust the structure of the planning flowchart appropriately according to the company's own conditions to achieve human-machine balance.

[0078] Example 2

[0079] On the basis of Example 1, the process link evaluation method based on graph neural network, step 1, comprises:

[0080] Step 11: Obtain the work data uploaded by the user, classify the work data to obtain the process data segment and execution data segment corresponding to this work, establish a matching relationship between the process data segment and the execution data segment, and generate several process links of this work;

[0081] Step 12: Analyze the interaction relationship between different process data segments, generate corresponding interaction conditions, match the corresponding process links for each interaction condition, use the interaction conditions to deploy data for the corresponding execution data segments, and use the data range corresponding to each process link to adjust the corresponding process link and generate a planning flowchart;

[0082] Step 13: using the graph neural network to identify the point representation vector corresponding to each process node in the planning flowchart, and respectively obtaining the node attributes corresponding to each process node, determining the interaction logic corresponding to the point representation vector according to the node attributes, and generating a plurality of interaction representation vectors of the planning flowchart;

[0083] Step 14: Determine the basic operator information of the corresponding process node according to the point representation vector, determine the deep operator information of the corresponding process node according to the interaction representation vector, match the corresponding operator elements for the process node according to the basic operator information, match the corresponding operator conditions for the operator elements according to the deep operator information, and generate a graph operator corresponding to each of the process nodes.

[0084] In this example, the process data segment represents data related to the workflow in the work data;

[0085] In this example, the execution data segment represents the work data and the data related to the execution work;

[0086] In this example, the process link represents a part of the process generated by combining an execution work and related data;

[0087] In this example, the point representation vector indicates the result of using a vector to represent the task in the process node;

[0088] In this example, the interaction representation vector is a vector generated by combining multiple point representation vectors with the same direction in the planning flowchart, and is used to represent a logical result;

[0089] In this example, the basic operator information includes the location and node tasks of the process node, and the deep operator information represents the interaction relationship between the process node and other process nodes;

[0090] In this example, the operator element represents the border corresponding to the process node, and the operator condition represents the condition required for the operator element to perform its work.

[0091] The working principle and beneficial effects of the above technical solution are as follows: by classifying the work data uploaded by the user and then matching it to generate several process links of this work, and then constructing interaction conditions based on the interaction relationship between the work data, and further matching the corresponding process links and data ranges for each interaction condition, thereby generating a planning flowchart, and then identifying the point representation vector corresponding to each process node in the planning flowchart and the interaction representation vectors between different process nodes in the graph neural network, thereby determining the basic operator information and deep operator information of each process node, and finally matching the corresponding operator elements and operator conditions for each process node, generating a graph operator for each process node, and laying the foundation for subsequent logical analysis.

[0092] Example 3

[0093] On the basis of Example 1, the process link evaluation method based on graph neural network, step 2, as Figure 2 As shown, including:

[0094] Step 21: generating a plurality of process directions of the planning flowchart according to the graph operator, dividing each of the process directions into a plurality of unit stages, respectively counting a plurality of unit stages corresponding to each process node, and generating a logic loop condition corresponding to each process node in the planning flowchart based on the process direction corresponding to each unit stage;

[0095] Step 22: Establish a logic model of the planning flowchart using the logic loop condition, input the work data into the logic model for logic analysis, generate a process result corresponding to each process trend, and obtain the data change process of the work data in the logic model;

[0096] Step 23: Based on the data change process and the process results corresponding to each process direction, determine the data processing characteristics and data processing results corresponding to each unit stage in the planning flowchart, generate node results of corresponding process nodes and display them.

[0097] In this example, the unit phase represents the process phase between two process nodes;

[0098] In this example, a process node can correspond to one or more logical loop conditions;

[0099] In this example, the data processing feature represents the feature corresponding to the data that needs to be processed in a unit stage;

[0100] In this example, the data processing result refers to the result generated after the data is input into the unit stage and processed.

[0101] The working principle and beneficial effects of the above technical solution: Since each process node has a corresponding processing flow, each process node needs to be analyzed and processed accordingly. First, the process direction of the planning flowchart is determined according to the graph operator, and then it is divided into several unit stages. The logical loop conditions of each process node of the planning flowchart are constructed, and the logical loop conditions are further used to establish the logical model of the planning flowchart. The process results of each process direction are determined through logical analysis, and the data change process of the working data in the logical model is determined. Finally, the data processing characteristics and data processing results corresponding to each node are determined, and the node results of each process node are generated. In this way, multiple process nodes can be analyzed at the same time. By using the model to simulate the working conditions of the process nodes, the actual calculation time can be reduced and the efficiency and quality of node analysis can be improved.

[0102] Example 4

[0103] On the basis of Example 3, the process link evaluation method based on graph neural network further includes:

[0104] Obtaining the work target characteristics uploaded by the user;

[0105] Filter the output node results corresponding to the output process nodes and generate process result features;

[0106] When the process result characteristics are inconsistent with the work target characteristics, the planning flowchart is determined to be abnormal, and an abnormal reminder is generated and displayed.

[0107] In this example, the output process node represents a process node for executing content output in the planning flow chart.

[0108] The working principle and beneficial effects of the above technical solution: By comparing the work target characteristics uploaded by the user with the process result characteristics of the output process node, it can be determined whether a large number of errors have occurred in the planning flowchart and corresponding reminders can be given.

[0109] Example 5

[0110] On the basis of Example 1, the process link evaluation method based on graph neural network, step 3, includes:

[0111] Step 31: input each of the node results into the graph neural network respectively, identify the local structure between different node results, generate an adjacency list of the planning flowchart in combination with the presentation data corresponding to each of the node results, and capture the process information corresponding to each of the process nodes in the adjacency list;

[0112] Step 32: Establishing the global structure of the planning flowchart according to the local structure between different node results, inputting the process information into the global structure, obtaining the planning role of each process node in the planning flowchart, and establishing the graph structure information of the corresponding process node according to the planning role;

[0113] Step 33: Determine the executable workload of the corresponding process node and the transferable workload between different process nodes according to the graph structure information, and determine the cost investment corresponding to each of the process nodes.

[0114] In this example, the local structure refers to a structure composed of two or more nodes;

[0115] In this example, the adjacency list represents a data structure used to express the relationship between different process nodes in the planning flow chart;

[0116] In this example, the global structure represents the structure composed of all process nodes in the planning flow chart;

[0117] In this example, the planning role represents the role of a process node in the planning flow chart;

[0118] In this example, the executable workload represents the maximum workload that can be carried when the process node executes the work;

[0119] In this example, the transferable workload represents the maximum workload for data transfer between process nodes.

[0120] The working principle and beneficial effects of the above technical solution are as follows: In order to perform cost analysis on each process node, an adjacency list is first established based on the local structure between the node results, and the process information corresponding to each process node is captured in the table. Then, the planning role of each process node in the planning flowchart is determined by analyzing the global structure of the planning flowchart, so as to establish corresponding graph structure information for the corresponding process node, and further analyze the executable workload and transferable workload of each process node, so as to determine the cost input of each process node. In this way, the structure, workload and structure of the planning flowchart of each process node can be analyzed, the cost of the planning flowchart can be allocated to each process node, and the cost input of each process node can be determined.

[0121] Example 6

[0122] Based on Example 1, the process link evaluation method based on graph neural network, step 4, includes:

[0123] Step 41: Map the graph structure information corresponding to each of the process nodes in the planning flowchart, add a corresponding external structure to each of the graph structure information according to the mapping result, and connect the graph structure information using the external structure to generate the overall graph structure information of the planning flowchart;

[0124] Step 42: traverse the overall graph structure information using the graph neural network to obtain a number of process decision statements of the planning flowchart, generate a decision tree of the planning flowchart, analyze the decision execution information corresponding to each leaf node in the decision tree, and screen and determine the overloaded leaf nodes with bottleneck information and the isolated leaf nodes with redundant information;

[0125] Step 44: Obtain the overload graph structure information corresponding to the overload leaf node, construct the execution process corresponding to the overload leaf node according to the overload graph structure information, select the corresponding experimental data from the work data and input it into the execution process to check whether the execution time of the corresponding overload leaf node is abnormal, and determine that the process node corresponding to the overload leaf node is a bottleneck process node;

[0126] Step 45: Obtain the isolated graph structure information corresponding to the isolated node, identify the input execution object and the output execution object corresponding to the isolated leaf node in the isolated graph structure information, and when the isolated leaf node does not contain an output execution object or does not contain an output execution object, determine that the process node corresponding to the isolated node is a redundant process node.

[0127] In this example, the external structure represents a graph structure that connects different process nodes;

[0128] In this example, the process decision statements represent the control statements corresponding to the execution of different tasks in the planning flow chart;

[0129] In this example, the process of checking whether the execution time of the corresponding overloaded leaf node is abnormal is: checking whether the execution time of the overloaded leaf node exceeds the average execution time of the planning flowchart.

[0130] The working principle and beneficial effects of the above technical solution are as follows: Since the planning flowchart may contain some unreasonable process nodes, the process decision statements contained in the planning flowchart are analyzed through the graph structure information to construct a decision tree, and the overloaded leaf nodes and isolated leaf nodes are further screened according to the decision execution information. The input-output relationship of the isolated leaf node is determined by analysis to further determine whether it is a redundant process node, and the overloaded leaf node is further determined by time screening to determine whether it is a bottleneck process node. In this way, the abnormal nodes in the planning flowchart can be quickly screened out, completing the basic work of the evaluation.

[0131] Example 7

[0132] On the basis of Example 6, the process of screening and determining overloaded leaf nodes with bottleneck information and isolated leaf nodes with redundant information in the process link evaluation method based on graph neural network includes:

[0133] Obtaining input-output relationships between different leaf nodes in the decision tree, and establishing a plurality of leaf node flows;

[0134] Respectively obtain the last leaf node corresponding to each of the leaf node streams, and add an isolated identifier to each of the last leaf nodes;

[0135] The node input-output duration corresponding to each leaf node flow and the mean square error duration corresponding to each leaf node flow are obtained respectively, the overloaded leaf nodes whose node input-output duration is higher than the corresponding mean square error duration are obtained, and a bottleneck identifier is added to each overloaded leaf node.

[0136] The working principle and beneficial effects of the above technical solution: Through the above method, leaf nodes with isolated characteristics and overloaded leaf nodes with timeout characteristics in the decision tree can be screened out, and corresponding labels can be added to them, thereby achieving the purpose of accurate identification and classification marking.

[0137] Example 8

[0138] On the basis of Example 1, the process link evaluation method based on graph neural network, the step 5 includes:

[0139] Step 51: respectively obtain the work pressure characteristics corresponding to each of the process nodes, use the work pressure characteristics to evaluate the rationality of the cost input corresponding to each of the process nodes, determine the first process node whose cost input is higher than the corresponding use cost, and the second process node whose cost input is lower than the corresponding use cost, and generate cost calibration suggestions;

[0140] Step 52: Locate each of the bottleneck process nodes and the redundant process nodes of the purpose one in the planning flowchart, obtain the pressure to be diverted corresponding to each of the bottleneck process nodes, and generate node modification suggestions in combination with deletion suggestions for each of the redundant process nodes.

[0141] The working principle and beneficial effects of the above technical solution are as follows: the rationality of each process node is evaluated separately, the process nodes with unreasonable costs are determined, and corresponding suggestions are configured for bottleneck process nodes and redundant process nodes. Cost calibration suggestions and node modification suggestions are generated for users' reference.

[0142] Example 9

[0143] Based on Example 1, the process link evaluation method based on graph neural network further includes:

[0144] A scoring report of the planning flowchart is generated and displayed according to the cost calibration suggestion and the node modification suggestion.

[0145] The working principle and beneficial effects of the above technical solution: After making suggestions, the planning flowchart is scored, allowing users to intuitively understand the defects of the planning flowchart.

[0146] Example 10

[0147] Based on Example 1, the process link evaluation method based on graph neural network further includes:

[0148] The graph neural network is used to evaluate several node functions corresponding to each of the process nodes, and a function scoring report for each of the process nodes is generated and displayed.

[0149] The working principle and beneficial effects of the above technical solution: Evaluating each process node separately can enable users to better understand the functions and defects of each node.

[0150] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A process link evaluation method based on graph neural network, characterized in that: include: Step 1: Build a planning flowchart for this work based on the work data uploaded by the user, and use the graph neural network to correct the graph operator corresponding to each process node in the planning flowchart; Step 2: Generate several logical loop conditions of the planning flow chart according to the graph operator, use the logical loop conditions to perform logical analysis on the work data, and obtain node results corresponding to process nodes; Step 3: Use the graph neural network to capture the process information contained in each of the node results, obtain the graph structure information corresponding to each of the process nodes, and analyze the cost input corresponding to each of the process nodes; Step 4: Establish the overall graph structure information of the planning flowchart, use the graph neural network to screen the bottleneck information and redundant information contained in the overall graph structure information, and determine the bottleneck process nodes and redundant process nodes of the planning flowchart; Step 5: Generate cost calibration suggestions based on the cost input corresponding to each of the process nodes, locate each of the bottleneck process nodes and each of the redundant process nodes in the planning flowchart, and generate node modification suggestions.

2. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: The step 1 comprises: Step 11: Obtain the work data uploaded by the user, classify the work data to obtain the process data segment and execution data segment corresponding to this work, establish a matching relationship between the process data segment and the execution data segment, and generate several process links of this work; Step 12: Analyze the interaction relationship between different process data segments, generate corresponding interaction conditions, match the corresponding process links for each interaction condition, use the interaction conditions to deploy data for the corresponding execution data segments, and use the data range corresponding to each process link to adjust the corresponding process link and generate a planning flowchart; Step 13: using the graph neural network to identify the point representation vector corresponding to each process node in the planning flowchart, and respectively obtaining the node attributes corresponding to each process node, determining the interaction logic corresponding to the point representation vector according to the node attributes, and generating a plurality of interaction representation vectors of the planning flowchart; Step 14: Determine the basic operator information of the corresponding process node according to the point representation vector, determine the deep operator information of the corresponding process node according to the interaction representation vector, match the corresponding operator elements for the process node according to the basic operator information, match the corresponding operator conditions for the operator elements according to the deep operator information, and generate a graph operator corresponding to each of the process nodes.

3. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: Step 2 includes: Step 21: generating a plurality of process directions of the planning flowchart according to the graph operator, dividing each of the process directions into a plurality of unit stages, respectively counting a plurality of unit stages corresponding to each process node, and generating a logic loop condition corresponding to each process node in the planning flowchart based on the process direction corresponding to each unit stage; Step 22: Establish a logic model of the planning flowchart using the logic loop condition, input the work data into the logic model for logic analysis, generate a process result corresponding to each process trend, and obtain the data change process of the work data in the logic model; Step 23: Based on the data change process and the process results corresponding to each process direction, determine the data processing characteristics and data processing results corresponding to each unit stage in the planning flowchart, generate node results of corresponding process nodes and display them.

4. A process link evaluation method based on graph neural network as claimed in claim 3, characterized in that: Also includes: Obtaining the work target characteristics uploaded by the user; Filter the output node results corresponding to the output process nodes and generate process result features; When the process result characteristics are inconsistent with the work target characteristics, the planning flowchart is determined to be abnormal, and an abnormal reminder is generated and displayed.

5. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: Step 3 includes: Step 31: input each of the node results into the graph neural network respectively, identify the local structure between different node results, generate an adjacency list of the planning flowchart in combination with the presentation data corresponding to each of the node results, and capture the process information corresponding to each of the process nodes in the adjacency list; Step 32: Establishing the global structure of the planning flowchart according to the local structure between different node results, inputting the process information into the global structure, obtaining the planning role of each process node in the planning flowchart, and establishing the graph structure information of the corresponding process node according to the planning role; Step 33: Determine the executable workload of the corresponding process node and the transferable workload between different process nodes according to the graph structure information, and determine the cost investment corresponding to each of the process nodes.

6. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: Step 4 includes: Step 41: Map the graph structure information corresponding to each of the process nodes in the planning flowchart, add a corresponding external structure to each of the graph structure information according to the mapping result, and connect the graph structure information using the external structure to generate the overall graph structure information of the planning flowchart; Step 42: traverse the overall graph structure information using the graph neural network to obtain a number of process decision statements of the planning flowchart, generate a decision tree of the planning flowchart, analyze the decision execution information corresponding to each leaf node in the decision tree, and screen and determine the overloaded leaf nodes with bottleneck information and the isolated leaf nodes with redundant information; Step 44: Obtain the overload graph structure information corresponding to the overload leaf node, construct the execution process corresponding to the overload leaf node according to the overload graph structure information, select the corresponding experimental data from the work data and input it into the execution process to check whether the execution time of the corresponding overload leaf node is abnormal, and determine that the process node corresponding to the overload leaf node is a bottleneck process node; Step 45: Obtain the isolated graph structure information corresponding to the isolated node, identify the input execution object and the output execution object corresponding to the isolated leaf node in the isolated graph structure information, and when the isolated leaf node does not contain an output execution object or does not contain an output execution object, determine that the process node corresponding to the isolated node is a redundant process node.

7. A process link evaluation method based on graph neural network as claimed in claim 6, characterized in that: The process of screening and determining overloaded leaf nodes with bottleneck information and isolated leaf nodes with redundant information includes: Obtaining input-output relationships between different leaf nodes in the decision tree, and establishing a plurality of leaf node flows; Respectively obtain the last leaf node corresponding to each of the leaf node streams, and add an isolated identifier to each of the last leaf nodes; The node input-output duration corresponding to each leaf node flow and the mean square error duration corresponding to each leaf node flow are obtained respectively, the overloaded leaf nodes whose node input-output duration is higher than the corresponding mean square error duration are obtained, and a bottleneck identifier is added to each overloaded leaf node.

8. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: The step 5 comprises: Step 51: respectively obtain the working pressure characteristics corresponding to each of the process nodes, use the working pressure characteristics to evaluate the rationality of the cost input corresponding to each of the process nodes, determine the first process node whose cost input is higher than the corresponding use cost, and the second process node whose cost input is lower than the corresponding use cost, and generate cost calibration suggestions; Step 52: Locate each of the bottleneck process nodes and the redundant process nodes of the purpose in the planning flowchart, obtain the pressure to be diverted corresponding to each of the bottleneck process nodes, and generate node modification suggestions in combination with deletion suggestions for each of the redundant process nodes.

9. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: Also includes: A scoring report of the planning flowchart is generated and displayed according to the cost calibration suggestion and the node modification suggestion.

10. A process link evaluation method based on graph neural network as claimed in claim 1, characterized in that: Also includes: The graph neural network is used to evaluate several node functions corresponding to each of the process nodes, and a function scoring report for each of the process nodes is generated and displayed.

Images