Digital one-network financial marketing management method and system

Abstract

The present application relates to the technical field of financial management, in particular to a digital one-network financial marketing management method and system, comprising the following steps: acquiring node data contrast queuing delay task sequence, reading dependent chain comparison time length analysis and splitting link extraction, extracting standby capacity contrast residual time weight distribution task, matching regional distribution order according to business currency time limit, reading branch comparison delay insertion position to obtain execution scheme. In the present application, the processing delay, the queuing quantity and the capacity of the transaction node are compared in real time, the task entering order and the node position are optimized, the dependent chain comparison and the link information extraction are combined, the split task is reorganized under the condition of maintaining the dependent order, the resource utilization rate is improved by matching the standby channel capacity, the task residual time and the dependent weight, the priority is adjusted according to the business type, the currency type and the time limit across regions, and the overall waiting time is shortened by inserting the branch of the confluence node according to the delay optimization.

Description

Technical Field

[0001] This invention relates to the field of financial management technology, and in particular to a digital, one-stop financial marketing management method and system. Background Technology

[0002] The field of financial management technology involves the organization, coordination, and implementation of core matters such as capital flow, asset allocation, risk control, return analysis, market transactions, and customer relations. It encompasses information collection, processing, and decision-making execution processes within banking, securities investment, insurance operations, tax processing, and other financial service systems. Its aim is to achieve efficient allocation and operation of financial resources through institutionalized, procedural, and tool-based methods. Traditional digital integrated financial marketing management methods refer to the centralized management and processing of marketing activities for multiple financial products and services on a unified digital platform. These methods typically rely on centralized databases to integrate multi-channel business information, utilize pre-set business logic rules for customer classification and business matching, and complete the execution and tracking of marketing tasks through batch automatic push and interactive recording.

[0003] Existing technologies lack continuous monitoring of node operating status and dynamic changes in task queues during multi-stage task processing. This leads to discrepancies between task allocation order and actual processing capacity, difficulty in reorganizing some tasks on the dependency chain, frequent idle capacity of backup processing channels, difficulty in simultaneously considering business urgency and time constraints in the sorting of cross-regional tasks, and uneven arrival rhythm of branches in merging nodes can cause cumulative delays, resulting in insufficient resource utilization and extended overall task response time. Summary of the Invention

[0004] To address the technical problems existing in the prior art, embodiments of the present invention provide a digital one-stop financial marketing management method, comprising the following steps:

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a digital one-stop financial marketing management method, comprising the following steps:

[0006] S1: Obtain transaction node processing latency, queue number and capacity data in the unified marketing scheduling platform, compare waiting time and processing latency, arrange task entry order and node position, and obtain node task execution sequence;

[0007] S2: Based on the node task execution sequence, read the task dependency chain data, compare the task execution time with the identifier sequence order, analyze the splittable positions, extract the link information, and obtain a list of split task chains;

[0008] S3: Based on the list of split task chains, extract the capacity of the backup channel in the transaction clearing channel, compare the remaining execution time of the task with the dependency weight, select the pushable tasks in order of capacity and allocate them to the target channel to obtain the backup channel push list;

[0009] S4: Based on the backup channel push list, according to the task business type, currency type and compliance time limit, associate the currency type and time limit with the business type, match the regulatory priority table with the region, allocate the processing order, and obtain the cross-regional priority processing list.

[0010] S5: Based on the cross-region priority processing list, read the number of branches of the merging node, compare the branch list with the cumulative delay value, filter out the branches that have not been reached and fill in the insertion position according to the delay value to obtain the execution path scheme of the merging node.

[0011] As a further aspect of the present invention, the node task execution sequence includes the task entry order, the node position to which the task belongs, and the comparison result of task waiting time and processing delay; the split task chain list includes the position of splittable tasks, task dependency link information, and sequence order comparison result; the backup channel push list includes the task remaining execution time sorting result, available backup channel capacity allocation result, and push task list; the cross-regional priority processing list includes the currency type associated with the business type, the compliance time limit corresponding to the currency type, and the regulatory priority of the region; and the merging node execution path scheme includes the unreached branch list, the insertion position corresponding to the cumulative delay value, and the required branch comparison result.

[0012] As a further aspect of the present invention, the task dependency chain data refers to reading the execution time and task identifier of each task based on the node task execution sequence, comparing them in sequence, extracting the predecessors between tasks, and arranging the dependency relationships in the original order.

[0013] The dependency weight refers to the weight value of the task in the dependency chain when comparing the backup channel capacity information with the remaining execution time of the task. The weight value comes from the original data record of the task dependency relationship and can be used as a task scheduling attribute to determine priority in subsequent steps in combination with the backup channel capacity and the remaining execution time.

[0014] As a further aspect of the present invention, the regulatory priority table refers to, after the business type compliance association table is established, matching the regional code in the regulatory priority table with the business type, merging the matched regional information into the task record, and grouping and sorting them according to the region.

[0015] The branch list refers to the process of reading the essential branch identifier sequence of the merging node after the cross-regional priority processing list is established, comparing the completed branches with the essential branches, filtering out the branches that have not been reached, and inserting them into the position according to the cumulative delay value.

[0016] As a further aspect of the present invention, the specific steps of S1 are as follows:

[0017] S101: Obtain real-time processing latency, current task queue count, and node capacity data of transaction nodes in the unified marketing scheduling platform. Partition the data by node according to the node identifier, and match the processing latency and task queue count within the partition with the node capacity to obtain node partition performance information.

[0018] S102: Based on the node partition performance information, according to the number of tasks queued for each node and the node capacity, and compared with the node processing latency, the comparison results are distributed in order of waiting time to obtain the node waiting time sequence.

[0019] S103: Based on the node waiting time sequence, match each task to the corresponding node position according to the order of entry, and connect the execution flow according to the position of the node in the sequence to obtain the node task execution sequence.

[0020] As a further aspect of the present invention, the specific steps of S2 are as follows:

[0021] S201: Based on the node task execution sequence, read the task execution dependency chain data, compare the task execution time with the task identifier in sequence, verify the position of each task in the sequence, and obtain the task order correspondence information;

[0022] S202: Based on the task order correspondence information, filter the task positions that can be split in the order, calculate the task position node occupancy rate, and list the task positions to obtain a list of splittable task positions;

[0023] S203: Based on the list of divisible task positions, extract the link information that matches the position in the list from the task execution dependency chain data, and connect them according to the sequential relationship of the tasks to obtain a list of split task chains.

[0024] As a further aspect of the present invention, the specific steps of S3 are as follows:

[0025] S301: Based on the split task chain list, extract the available capacity of the backup channel buffer in the financial transaction clearing channel, partition the capacity data according to the channel identifier, and obtain the backup channel capacity information by corresponding capacity and task position.

[0026] S302: Based on the backup channel capacity information, compare the remaining execution time of each task with the task dependency weight, calculate the task dependency weight occupancy ratio, and arrange the tasks in order according to the remaining execution time to obtain the task execution order table;

[0027] S303: Based on the task execution priority table, select the tasks that can be pushed to the backup channel in sequence, assign the tasks to the corresponding target channels, and obtain the backup channel push list.

[0028] As a further aspect of the present invention, the specific steps of S4 are as follows:

[0029] S401: Based on the backup channel push list, read the transaction task business type, transaction currency type and compliance time limit, match the transaction currency type and compliance time limit according to the business type, and link the association relationship according to the task identifier to obtain the business type compliance association table.

[0030] S402: Based on the business type compliance association table, match the region to which the business type belongs in the regulatory priority table, merge the matched region code with the original association information, and obtain the region matching result table;

[0031] S403: Based on the aforementioned regional matching result table, the task processing priority is arranged according to the compliance time limit, and the tasks are assigned to the corresponding regions to obtain a cross-regional priority processing list.

[0032] As a further aspect of the present invention, the specific steps of S5 are as follows:

[0033] S501: Based on the cross-regional priority processing list, read the number of branches completed by the merging node, match the number with the required branch list parameter of the merging node, identify the branch arrival status, and obtain the branch arrival status table.

[0034] S502: Based on the branch arrival status table, compare the unreached branches with the cumulative delay value parameter, and distribute the branches according to the cumulative delay value in the list to obtain the unreached branch delay position table.

[0035] S503: Based on the unreached branch delay position table, insert the branch into the corresponding position in the required branch list parameter of the merging node, and connect them into an executable node path to obtain the merging node execution path scheme.

[0036] The digital, all-in-one financial marketing management system includes:

[0037] The transaction node monitoring module obtains the real-time processing latency, task queue number and processing capacity of transaction nodes in the unified marketing scheduling platform. It groups the queue number and processing capacity by node, compares the waiting time and processing latency, and assigns the order of task entry and the position of the node according to the comparison results to obtain the node task execution sequence.

[0038] The task chain decomposition module reads the task dependency chain data and execution time based on the node task execution sequence, compares the execution time with the order, identifies the separable task positions, extracts the corresponding links, and obtains a list of split task chains.

[0039] Based on the list of split task chains, the backup channel allocation module extracts the backup channel capacity of the financial transaction clearing channel, compares the remaining execution time of the task with the dependency weight, and selects the pushable tasks in order of capacity to allocate them to the target channel, thus obtaining the backup channel push list.

[0040] The cross-regional priority processing module pushes a list based on the backup channel, and associates the currency type and time limit with the business type according to the transaction task business type, currency type and compliance time limit, matches the regulatory priority table with the region, allocates the processing order, and obtains the cross-regional priority processing list.

[0041] Based on the cross-regional priority processing list, the merging path planning module calls the number of completed branches, the list of required branches, and the cumulative delay value of the merging node. It compares the branch list with the cumulative delay value, filters out branches that have not been reached, and adds them to the positions according to the delay value to obtain the execution path scheme of the merging node.

[0042] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0043] In this invention, by comparing the processing latency, queue size and capacity of transaction nodes in real time, the order of task entry and node position are optimized. Combined with dependency chain comparison and link information extraction, divisible tasks can be recombined while maintaining the dependency order. The resource utilization rate is improved by matching the capacity of the backup channel with the remaining time of the task and the dependency weight. Priority is adjusted across regions according to business type, currency type and time limit. Merging nodes optimize branch insertion based on latency to shorten the overall waiting time. Attached Figure Description

[0044] 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.

[0045] Figure 1 This is a schematic diagram of the steps of the present invention;

[0046] Figure 2 This is a detailed schematic diagram of S1 of the present invention;

[0047] Figure 3 This is a detailed schematic diagram of S2 of the present invention;

[0048] Figure 4 This is a detailed schematic diagram of S3 of the present invention;

[0049] Figure 5 This is a detailed schematic diagram of S4 of the present invention;

[0050] Figure 6 This is a detailed schematic diagram of S5 of the present invention;

[0051] Figure 7 This is a system module diagram of the present invention. Detailed Implementation

[0052] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0053] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0054] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0055] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0056] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0057] Please see Figure 1 This invention provides a digital, unified financial marketing management method, comprising the following steps:

[0058] S1: Obtain real-time processing latency, current task queue count, and node capacity data of transaction nodes in the unified marketing scheduling platform. After grouping by node identifier, compare the waiting time and processing latency item by item. Based on the comparison results, arrange the entry order and node position of each task to be processed to obtain the node task execution sequence.

[0059] S2: Based on the node task execution sequence, read the task execution dependency chain data, compare the task execution time and task identifier with the sequence order, analyze the splittable task positions, extract the corresponding link information, and obtain a list of split task chains.

[0060] S3: Based on the list of split task chains, extract the available capacity of the backup channel buffer in the financial transaction clearing channel, compare the remaining execution time of the tasks with the task dependency weight, sort the tasks from shortest to longest remaining execution time, select the pushable tasks in sequence according to the backup capacity, and allocate them to the target channel to obtain the backup channel push list.

[0061] S4: Based on the backup channel push list, according to the transaction task business type, transaction currency type and compliance time limit, the transaction currency type and compliance time limit are associated with the corresponding business type, matched with the region in the regulatory priority table, and then the processing order of tasks is allocated according to the compliance time limit to obtain the cross-regional priority processing list.

[0062] S5: Based on the cross-region priority processing list, read the number of branches completed by the merging node, compare the required branch list parameter of the merging node with the cumulative delay value parameter item by item, filter out the branches that have not been reached, and supplement the insertion position according to the cumulative delay value parameter to obtain the execution path scheme of the merging node.

[0063] The node task execution sequence includes the task entry order, the node position of the task, and the comparison results of task waiting time and processing delay. The split task chain list includes the position of splittable tasks, task dependency link information, and sequence order comparison results. The backup channel push list includes the sorting results of the remaining execution time of tasks, the allocation results of available backup channel capacity, and the list of push tasks. The cross-regional priority processing list includes the currency type associated with the business type, the compliance time limit corresponding to the currency type, and the regulatory priority of the region. The merge node execution path scheme includes the list of branches not yet reached, the insertion position corresponding to the cumulative delay value, and the comparison results of the required branches.

[0064] Please see Figure 2 The specific steps of S1 are as follows:

[0065] S101: Obtain real-time processing latency, current task queue count, and node capacity data of transaction nodes in the unified marketing scheduling platform. Partition the data by node according to the node identifier, and match the processing latency and task queue count within the partition with the node capacity to obtain node partition performance information.

[0066] First, the start and end times of each node's processing are retrieved. The difference between the end and start times is converted into millisecond-level latency and recorded as the processing latency value. For example, if node A processes a task with a start time of 08:00:00 and an end time of 08:00:05, the processing latency is 5000 milliseconds. The current number of tasks in the queue is obtained from the node's queue length field and directly regarded as the number of tasks in the queue. For example, if node A has 12 tasks in the queue at a certain moment, the number of tasks in the queue is recorded as 12. Node capacity data is read from the node's maximum concurrent processing capacity field. If node A can process a maximum of 15 tasks simultaneously, the node capacity is 15. After completing these three types of data collection, all data is partitioned according to the node identifier. Data with the same node identifier is grouped into the same interval. For example, the processing latency, queue count, and capacity of node A are all assigned to the partition of node A. Then, within the partition, the processing latency and queue count are mapped according to the node capacity. Specifically, a one-to-one matching relationship is established within the same partition. The processing latency and queue count of a node are compared with the node's capacity value. For example, if node A has a processing latency of 5000 milliseconds, a queue count of 12, and a capacity of 15, then the ratio of the number of tasks to the capacity is 12 / 15. The processing latency corresponds to a capacity unit of 15 for 5000 milliseconds. This process is repeated for all nodes. The processing latency and queue count in each node partition are mapped according to the capacity, and finally, the node partition performance information is obtained.

[0067] S102: Based on the node partition performance information, according to the number of tasks queued for each node and the node capacity, and compared with the node processing latency, the comparison results are distributed in order of waiting time to obtain the node waiting time sequence.

[0068] First, extract two data items for each node: the number of tasks in the queue and the node capacity. To map the number of tasks in the queue to the capacity, we need to read the total number of tasks currently queued for that node and the maximum number of tasks that node can process simultaneously. For example, if node B currently has 20 tasks in the queue and a maximum processing capacity of 25, then the task-to-capacity ratio for that node is 20 to 25. This ratio is then compared with the node's processing latency. The processing latency is obtained based on existing latency data in the node's partition performance information. For example, if node B's average processing latency is 7500 milliseconds, then the mapping between "20 to 25" and "7500 milliseconds" for that node can be directly established. This mapping applies to all nodes. After processing, a set of corresponding node ratios and processing delays is formed. These sets are then distributed according to waiting time, that is, the processing delay values ​​are placed in the sequence from shortest to longest, while retaining the node task ratio information corresponding to that delay. For example, if the ratio of node C is 10:15 and the processing delay is 4000 milliseconds, and the ratio of node D is 18:20 and the processing delay is 9000 milliseconds, then in the distribution process, node C is placed before node B, and node B is placed before node D. This distribution process determines the position of all nodes according to the order of their corresponding waiting times and includes them in the sorting result along with the task ratio, finally obtaining the node waiting time sequence.

[0069] S103: Based on the node waiting time sequence, match each task to the corresponding node position according to the order of entry, and connect the execution flow according to the position of the node in the sequence to obtain the node task execution sequence;

[0070] First, the entry sequence number in each task record is read sequentially. This number indicates the order in which tasks arrive. For example, if task T01 is numbered 1, T02 is numbered 2, and T03 is numbered 3, it means T01 entered the processing stage before T02 and T03. Then, the task numbers are matched with the corresponding node identifiers in the node waiting time sequence. During this matching process, the node affiliation of each task needs to be checked item by item to ensure that the correspondences, such as task T01 belonging to node N1, task T02 belonging to node N3, and task T03 belonging to node N2, are accurate. Next, according to the position order of the nodes in the waiting time sequence, the matched task information is connected to form an execution chain. For example, the waiting time for node N2 in the waiting time sequence is... With a time interval of 4500 milliseconds, 6000 milliseconds for node N1, and 8000 milliseconds for node N3, the task at node N2 will be processed before the tasks at nodes N1 and N3 in the execution chain. Subsequently, tasks are filled in sequentially according to the chain structure. For example, T03 falls into N2, T01 falls into N1, and T02 falls into N3, forming a task sequence that is connected according to the node positions. If there are multiple tasks under the same node, they will continue to be connected at that node position according to the entry order. For example, if the task number T01 under node N1 is less than the task number T05, then T01 comes first and T05 comes last. Through this connection process, the entry order of tasks and the processing order of nodes are completely integrated to obtain the node task execution sequence.

[0071] Please see Figure 3 The specific steps of S2 are as follows:

[0072] S201: Based on the node task execution sequence, read the task execution dependency chain data, compare the task execution time with the task identifier in sequence, verify the position of each task in the sequence, and obtain the task sequence information.

[0073] First, read the task execution dependency chain data line by line, extracting the execution duration of each task in the dependency chain along with its corresponding task identifier. For example, task T11 has an execution duration of 6200 milliseconds and its task identifier is T11, while task T15 has an execution duration of 4800 milliseconds and its task identifier is T15. Arrange these task data sequentially and compare them with the corresponding task order in the node's task execution sequence. The comparison process requires checking whether the position of each task in the execution sequence matches the order recorded in the dependency chain. If task T15 is after T11 in the dependency chain, it must also be after T11 in the execution sequence. If any inconsistency is found, further comparison is needed. The difference position is directly marked and the comparison continues to the next task. The whole process needs to be repeated until all tasks are compared. During this period, for multiple tasks under the same node, it is also necessary to compare whether their order of records in the dependency chain is consistent with their relative positions in the execution sequence. For example, T21 and T23 under node N4, T21 precedes T23 in the dependency chain, so T21 must also precede T23 in the execution sequence. This position comparison is performed not only between different nodes, but also within the same node. In this way, the duration and identification information of the dependency chain data are fully checked into the order information of the node task execution sequence, and finally the task order correspondence information is obtained.

[0074] S202: Based on the task priority information, filter the task positions that can be split in the priority, calculate the task position node occupancy rate, and add the task positions to the list to obtain the list of splittable task positions;

[0075] The specific formula for calculating the occupancy rate of task location nodes is as follows:

[0076] ;

[0077] Among them, This represents the occupancy rate of the task location node at time t. This represents the number of tasks assigned at the i-th task position at time t. This represents the maximum number of tasks that the i-th task location node can accommodate. This represents the number of tasks completed at the i-th task position node at time t. The critical weight coefficient represents the i-th task location node, and n represents the total number of task location nodes;

[0078] The formula's calculation logic is as follows: First, within each node, obtain the difference between the number of assigned tasks and the node's maximum capacity, and take the absolute value to eliminate the influence of positive and negative directions. Divide this difference by the node's maximum capacity to obtain a relative proportion. Then, multiply this proportion by the corresponding node's weight value to reflect the node's importance in the overall system. Sum the weighted proportions of all nodes to obtain the global capacity difference weighted sum. Next, within each node, obtain the difference between the number of assigned tasks and the number of completed tasks. Divide this difference by the node's maximum capacity to obtain the proportion of uncompleted tasks in the capacity. Square this proportion to amplify the impact of larger deviations, and sum the squared results of all nodes to obtain the global task progress difference sum. Then, take the square root of this sum to adjust the dimensions back to a level consistent with the proportion. Finally, add the aforementioned capacity difference weighted sum to the square root of the task progress difference squared sum to obtain the comprehensive occupancy value. Normalize this comprehensive value by using the sum of all node weight values ​​as the divisor, thus obtaining the task location node occupancy rate used to measure the overall task allocation and progress status.

[0079] Through monitoring and retrieval of scheduling task data:

[0080] Node 1: Maximum Capacity =120 tasks have been assigned =100 tasks completed =90 tasks, weight =0.8;

[0081] Node 2: Maximum Capacity =150 tasks have been assigned =130 tasks completed =110 tasks, weight =0.6;

[0082] Node 3: Maximum Capacity =100 tasks have been assigned =95 tasks completed =85 tasks, weight =0.9;

[0083] Node capacity was measured through stress testing. The number of allocations and completions was directly calculated from the scheduling logs. The weights were quantized and normalized to the 0–1 range by the percentage of impact.

[0084] Substitute into the calculation process:

[0085] Calculate the weighted term for the absolute difference proportion:

[0086] Node 1: ;

[0087] Node 2: ;

[0088] Node 3: ;

[0089] Weighted sum = 0.10664 + 0.07998 + 0.045 = 0.23162;

[0090] Calculate the standard deviation of the difference between allocation and completion:

[0091] Node 1: (100-90) / 120=0.08333, squared=0.00694;

[0092] Node 2: (130-110) / 150=0.13333, squared=0.01778;

[0093] Node 3: (95-85) / 100=0.1, squared=0.01;

[0094] Sum of squares = 0.00694 + 0.01778 + 0.01 = 0.03472, Root of square = 0.18626;

[0095] Calculate the weighted sum of the denominators:

[0096] ;

[0097] Substitute into the formula:

[0098] ;

[0099] Interpretation of results and numerical significance: The calculated task location node occupancy rate is 0.18169, which indicates that the current task location node has a low occupancy rate under the maximum task capacity, meaning that tasks can be added to the "splitable task location list".

[0100] Explanation of the innovative aspects of the formula:

[0101] The advantage of the formula is that by combining the weighted difference ratio of task capacity with the dispersion term of task completion and allocation differences, it can simultaneously reflect capacity deviation and schedule balance in the calculation of occupancy rate. Furthermore, by normalizing the weights, it eliminates the impact of differences in the size of differentiated nodes, making the calculation results usable for global scheduling evaluation.

[0102] S203: Based on the list of splittable task positions, extract the link information that is consistent with the list position from the task execution dependency chain data, and connect them according to the sequential relationship of the tasks to obtain a list of split task chains.

[0103] First, the task identifier information corresponding to each position in the list is read. For example, position 1 corresponds to task T08, position 2 corresponds to task T12, and position 3 corresponds to task T19. Then, the task execution dependency chain data is searched line by line. The task identifiers in the dependency chain are compared with the task identifiers in the list. If the task identifier recorded in the dependency chain is completely consistent with an identifier in the list, the complete link information of that task is extracted. The link information must include the predecessor task identifier, the successor task identifier, and the dependency relationship parameters associated with it. For example, if task T08 is located after T05 and before T10 in the dependency chain, then the association field and sequence information of T05 and T10 are obtained simultaneously during extraction. After extracting the link information of all matching tasks, the link information is connected according to the order of the tasks in their original execution sequence. During the connection, the dependency relationship between tasks must be maintained. For example, if T08 was originally located between T05 and T10 in the sequence, it must still be located between T05 and T10 when the chain is connected. If multiple tasks come from the same node, they are connected in the original order of entry at that node. For example, if task T12 enters before task T19 in node N7, then T12 is ranked before T19 in the chain. After the connection processing of all task link information, a continuous path containing all splittable task links is formed, and finally a list of split task chains is obtained.

[0104] Please see Figure 4 The specific steps of S3 are as follows:

[0105] S301: Based on the list of split task chains, extract the available capacity of the backup channel buffer in the financial transaction clearing channel, partition the capacity data according to the channel identifier, and obtain the backup channel capacity information by corresponding capacity and task position.

[0106] First, the identifier information and channel number of each task in the chain are read sequentially. The financial transaction clearing channel where the task is located is used as the positioning basis. The available capacity data is read from the backup channel buffer of the corresponding channel. The available capacity is obtained directly by calling the remaining task capacity field of the buffer. For example, the remaining task capacity of backup channel C1 is 120, and that of backup channel C2 is 85. All the obtained backup channel capacity values ​​are matched one-to-one with the channel identifiers to form a paired dataset of capacity and channel identifiers. Then, this paired data is partitioned according to the channel identifier, that is, capacity datasets with the same identifier are grouped into the same segment. For example, all capacity records under channel C1 are grouped into... In the same segment, channel C2 is processed in the same way. After partitioning, the partitioned capacity data needs to be matched with the position of the corresponding task in the split task chain list. For each task, the capacity value in its channel partition is found. For example, if task T52 in the chain belongs to channel C1, the capacity value in partition C1 is matched with T52 and recorded in the matching result. If there are multiple tasks under the same channel in the chain, the current available capacity data of the channel is assigned to each task when matching, until all tasks in the chain are associated with the capacity information. In this way, each task not only retains its position in the chain, but also has the associated information of available capacity, and finally the spare channel capacity information is obtained.

[0107] S302: Based on the backup channel capacity information, compare the remaining execution time of each task with the task dependency weight, calculate the task dependency weight occupancy ratio, and arrange the tasks in order of remaining execution time to obtain the task execution order table;

[0108] The specific formula for calculating the proportion of task dependency weight is as follows: ;

[0109] in, This represents the proportion of the task dependency weights for the v-th task. This represents the task dependency weight of the v-th task. This represents the arithmetic mean of the weights that all tasks depend on. This represents the remaining execution time of the v-th task. The arithmetic mean of the remaining execution times of all tasks is represented by , and u represents the total number of tasks currently pending execution. This represents the task dependency weight of the j-th task;

[0110] Formula operation logic: The dimensionless difference between the current task dependency weight and the average of all task dependency weights is obtained by subtraction and the absolute value is taken to remove directionality and keep the difference non-negative. Then, the ratio of the remaining execution time of the current task to the average remaining execution time of all tasks is calculated and the square root is calculated to form a dimensionless scaling factor for the time scale. The above difference is multiplied and combined with the time scaling factor and then multiplied with the backup channel capacity normalization coefficient to introduce the influence of the current available channel capacity. Finally, the average of all task dependency weights is used for division normalization to keep the result within a comparable dimensionless range, thus obtaining the task dependency weight occupancy ratio value that can be directly used for task priority ranking.

[0111] Monitor the rated bandwidth of the link. ;

[0112] Average utilization rate within the monitoring window ,calculate ;

[0113] and :calculate average batch ;

[0114] Example data: , ;

[0115] have to T=2.5h;

[0116] Number of downstream tasks ;

[0117] Schedule Standard Task slack ;

[0118] Calculate relaxation gap ;

[0119] Resource blocking ratio ;

[0120] Maximum value: , , ;

[0121] Normalization yields ;

[0122] ;

[0123] ;

[0124] Addition and subtraction The interval scaling (minimum 1.1587, maximum 3.0000) yields... ;

[0125] Calculated ;

[0126] Substituting into the formula for calculation (mean of the denominator) , ):

[0127] v=1: , ;

[0128] v=2: , ;

[0129] v=3: , ;

[0130] v=4: , ;

[0131] The shift schedule is based on the company's work system;

[0132] This result indicates that: Tasks 3 and 2 have a higher dependency weight in this batch, which is used as the sorting weight, along with the remaining execution time. A joint decision-making priority list for task execution;

[0133] Explanation of the innovative aspects of the formula:

[0134] The advantage of the formula lies in its reliance on the deviation term. The square root of the time ratio Normalization factor of spare channel capacity The same-dimensional synthesis can comprehensively reflect the impact of dependency intensity dispersion, relative remaining time scale, and available link capacity ratio on scheduling priority at the same numerical scale, resulting in... Values ​​can directly drive task sorting.

[0135] S303: Based on the task execution priority table, select the tasks that can be pushed to the backup channel in sequence, assign the tasks to the corresponding target channels, and obtain the backup channel push list;

[0136] First, starting from the first task in the priority table, read the task records sequentially. Each task record contains a task identifier, its owning node information, and a current pushable status flag. Tasks with a pushable status flag of "Yes" need to enter the backup channel allocation stage. For example, if task T07 in the priority table has a pushable status of "Yes," then T07 is treated as a task to be allocated. Its owning node number is read, and the backup channel identifier corresponding to that node number is searched in the backup channel capacity information. After finding an available channel, a correspondence is established between task T07 and that channel, and an allocation record is generated. Continue reading the next task in the priority table. If T12 is also pushable and belongs to the same backup channel, then the allocation record for that task is appended to that channel. If a task is found to be "no" in the pushable status, it is skipped and the process continues to read the sequence without participating in the allocation. This continues until all tasks in the sequence table have been checked. During the process, if the available capacity of the backup channel is less than the number of tasks to be allocated, only the number of tasks within the capacity limit in the sequence table will be allocated. For example, if the remaining capacity of backup channel C5 is 3 tasks, and the tasks that meet the conditions in the sequence table are T15, T16, T19, and T20, then only tasks T15, T16, and T19 will be allocated to C5, while T20 will be kept in the sequence table to wait for the next round of allocation. Through this one-by-one matching and allocation operation, all pushable tasks in the sequence table will be assigned to the corresponding backup channels, and the backup channel push list will be obtained.

[0137] Please see Figure 5 The specific steps of S4 are as follows:

[0138] S401: Based on the backup channel push list, read the transaction task business type, transaction currency type and compliance time limit, match the transaction currency type and compliance time limit according to the business type, and link the association relationship according to the task identifier to obtain the business type compliance association table.

[0139] First, the task identifier for each transaction task in the list is read sequentially. Then, the corresponding business type, currency type, and compliance time limit fields are retrieved from the transaction task information database. For example, task T21 has a business type of B1, a currency type of USD, and a compliance time limit of 48 hours. When mapping the currency type and compliance time limit according to the business type, all tasks of the same business type need to be categorized and their currency types and compliance time limits are paired to form records. For example, tasks under type B1 may have multiple sets of data such as USD-48 hours and EUR-72 hours. These paired records will be concatenated with the task identifier later. The concatenation process is to establish a one-to-one mapping relationship between each paired record and its corresponding task identifier. For example, USD-48 hours corresponds to T21, and EUR-72 hours corresponds to T26. This ensures that each task identifier forms a complete data combination with its business type, currency type, and compliance time limit. If multiple tasks have the same business type and currency type but different compliance time limits during processing, they are distinguished according to the value of the compliance time limit, and all different pairing results are retained in the mapping. For example, JPY-24 hours and JPY-36 hours under type B2 will be bound to T32 and T35 respectively. The entire process will repeat this operation for all tasks in the backup channel push list until all task identifiers have completed the business type correspondence and connection of currency type and compliance time limit, and finally obtain the business type compliance association table.

[0140] S402: Based on the business type compliance association table, match the region to which the business type belongs in the regulatory priority table, merge the matched region code with the original association information, and obtain the region matching result table;

[0141] First, the business type field of each record in the table is read. This field identifies the business category to which the transaction task belongs. For example, the business type of task T31 is B3, and the business type of task T42 is B7. Then, the regulatory priority table is searched for records with the same business type. During the search, the complete consistency of the business type code must be compared to ensure that there are no matching errors due to similar characters. For example, B3 can only match B3, not B30 or B13. When a match is successful, the region code corresponding to the business type in the regulatory priority table is read. For example, the region code matched by B3 is R2, and the region code matched by B7 is R5. This region code is added to the record of the corresponding task in the original business type compliance association table, while keeping the original transaction currency type and compliance time limit information in the record unchanged. For example, the original record for task T31 was B3, USD-48 hours, and after merging, it was updated to B3, USD-48 hours, R2. The original record for task T42 was B7, JPY-72 hours, and after merging, it was updated to B7, JPY-72 hours, R5. After adding the region codes for all tasks, it is necessary to check again whether there are cases where the same business type appears in multiple regions. If such a case exists, each region information is kept as an independent record in the result table without merging. For example, when B4 has a match in both R1 and R3, the corresponding task record will be split into two, keeping the records in regions R1 and R3 respectively. After adding the region codes, all task records form a complete dataset with business type, transaction currency type, compliance time limit, and region code, ultimately resulting in the region matching result table.

[0142] S403: Based on the regional matching result table, the task processing priority is arranged according to the compliance deadline, and the tasks are assigned to the corresponding regions to obtain a cross-regional priority processing list;

[0143] First, extract the compliance time limit and region code for each task record in the table. The compliance time limit should be read directly from the task record. For example, task T51 has a compliance time limit of 36 hours and belongs to region R3, while task T57 has a compliance time limit of 72 hours and belongs to region R2. Group all tasks by region code, and within each region group, sort them according to the length of the compliance time limit, with tasks with shorter time limits listed first. For example, in region R3, T51 (36 hours) should be listed before T64 (60 hours). After sorting within each region, use the sorting results as the task processing priority table for that region, while retaining the original business categories. The task list is generated by taking into account factors such as the type of transaction currency and other related information. Then, according to the priority table for each region, the tasks are sequentially assigned to the task execution list of the corresponding region. During the assignment, it is necessary to ensure that a task only enters the list of its own region and does not overlap with other regions. For example, T51 can only enter the execution list of R3 and will not be put into the list of R2 or R5. For multiple tasks with the same compliance time limit, such as T57 and T60 in R2, which both have a compliance time limit of 72 hours, their order of appearance in the original region's matching result table is used to arrange their positions. This assignment process will be executed sequentially in all regions until all tasks have been sorted and assigned, and finally, a cross-regional priority processing list is obtained.

[0144] Please see Figure 6 The specific steps of S5 are as follows:

[0145] S501: Based on the cross-regional priority processing list, read the parameter of the number of branches completed by the merging node, match the number with the parameter of the required branch list of the merging node, identify the branch arrival status, and obtain the branch arrival status table.

[0146] First, read each task record in the list that involves merging nodes to determine the merging node number to which these tasks belong. For example, tasks T72 and T75 both belong to merging node M4. Next, read the number of completed branches parameter for that node from the node status dataset. This parameter is an integer value used to represent the number of branches that have been successfully reached on that node. For example, M4 currently has 3 completed branches. Match this number with the required branch list parameter for the same node. The required branch list parameter contains all the branch identifiers required for the node to complete the merging. For example, M4's required branch list is [P1, P2, P3, P4, P5]. During the matching process, the completed branches need to be compared one by one. The identification of a branch is checked against the identification in the list of required branches. If they match, the branch is considered to have reached its destination; otherwise, it is considered to have not reached its destination. For example, if the identification of a completed branch is [P1, P3, P4], then P1, P3, and P4 are considered to have reached their destinations, while P2 and P5 are considered to have not reached their destinations. During the identification process, the status of each branch needs to be recorded separately and accompanied by its corresponding node number so as to maintain consistency with the node correspondence in subsequent steps. After the correspondence between the number of required branches and the number of completed branches is checked, a complete list of branch statuses can be formed, which clearly indicates which branches have reached their destinations and which branches have not yet reached their destinations, thus obtaining the branch arrival status table.

[0147] S502: Based on the branch arrival status table, compare the unreached branches with the cumulative delay value parameter, and distribute the branches according to the cumulative delay value in the list to obtain the unreached branch delay position table.

[0148] First, records of branches in the "not arrived" status are filtered out. These records must retain the branch identifier and the merging node number to which they belong. For example, the not arrived branches of node M4 are P2 and P5, and the not arrived branches of node M6 are P8 and P9. Next, the cumulative delay value parameter of each not arrived branch in the delay information table is read. This parameter represents the cumulative delay time of the branch in numerical form. For example, the cumulative delay value of P2 is 2400 milliseconds, P5 is 4800 milliseconds, P8 is 1500 milliseconds, and P9 is 3000 milliseconds. After obtaining the cumulative delay value, the not arrived branches under the same node are arranged in the node's delay position list according to their cumulative delay value, with shorter delay times listed first. The longer delay is listed last. For example, in node M4, if the cumulative delay value of P2 is less than that of P5, then P2 comes first and P5 comes last. In node M6, P8 comes first and P9 comes last. When crossing nodes, the delay position list is generated independently to ensure that the delay position distribution of each node does not affect each other. During the processing, if there are cases where the cumulative delay values ​​are the same, their original order in the branch arrival status table is kept unchanged. This can avoid disorderly changes to the relative relationship of the original data. After the delay position distribution of all unarrived branches of all nodes is completed, these results are integrated into a unified format, recording the order and cumulative delay value of each unarrived branch under each node in the delay position list, and finally obtaining the unarrived branch delay position table.

[0149] S503: Based on the unreached branch delay position table, insert the branch into the corresponding position in the required branch list parameter of the merging node, and connect them into an executable node path to obtain the merging node execution path scheme;

[0150] First, the branch identifier of each unreached branch and its ranking in the delay position list are read one by one. For example, P2 of node M4 is in position 1 and P5 is in position 2 of the delay list, and P8 of node M6 is in position 1 and P9 is in position 2. Then, the required branch list parameters for the merging nodes are extracted. This list records all the branch positions required for each merging node to complete execution and their fixed order. For example, the required branch list for node M4 is P1, P2, P3, P4, P5, and the required branch list for node M6 is P6, P7, P8, P9, P10. After extraction, the unreached branches are inserted one by one into the specified positions in the corresponding required branch list. This process requires matching the branch identifier with each item in the required branch list. After successful matching, insert the branch record from the delay position table at that position, while maintaining the relative order of the branches that have been reached in the original required branches. For example, in node M4, P2 should be inserted between P1 and P3, and P5 should be inserted after P4. In node M6, P8 should be inserted between P7 and P9, and P9 should be inserted after P8. After the insertion is completed, check the connection between each branch to ensure that the predecessor and successor of each branch are continuous in the logical path. For example, the successor of P1 is P2, the successor of P2 is P3, the successor of P3 is P4, and the successor of P4 is P5. Finally, a directly executable branch path chain is formed within each node. The branch chains of each node are combined in sequence according to the node identifier to obtain the execution path scheme of the merging node.

[0151] Please see Figure 7 The digital all-in-one financial marketing management system includes:

[0152] The transaction node monitoring module obtains the real-time processing latency, task queue number and processing capacity of transaction nodes in the unified marketing scheduling platform. It groups the queue number and processing capacity by node, compares the waiting time and processing latency, and assigns the order of task entry and the position of the node according to the comparison results to obtain the node task execution sequence.

[0153] The task chain decomposition module reads task dependency chain data and execution time based on the node task execution sequence, compares the execution time with the order, identifies separable task positions, extracts the corresponding chain, and obtains a list of split task chains.

[0154] The backup channel allocation module extracts the backup channel capacity of the financial transaction clearing channel based on the split task chain list, compares the remaining execution time of the task with the dependency weight, and selects the pushable tasks in order of capacity to allocate them to the target channel, thus obtaining the backup channel push list.

[0155] The cross-regional priority processing module pushes a list based on the backup channel. It associates the currency type and time limit with the business type according to the transaction task business type, currency type and compliance time limit, matches the regulatory priority table with the region, allocates the processing order, and obtains the cross-regional priority processing list.

[0156] The merging path planning module is based on the cross-regional priority processing list. It calls the number of completed branches, the list of required branches, and the cumulative delay value of the merging node. It compares the branch list with the cumulative delay value, filters out branches that have not been reached, and adds them to the positions according to the delay value to obtain the execution path plan of the merging node.

[0157] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for digitizing and managing financial marketing of a network, characterized by, Comprise the following steps: S1: Obtain the transaction node processing delay, queue quantity and capacity data in the unified marketing scheduling platform, compare the waiting time and processing delay, arrange the task entering sequence and node position, and obtain the node task execution sequence; S2: Based on the node task execution sequence, read the task dependency chain data, compare the task execution time and sequence order, analyze the splittable position, extract the link information, and obtain the split task chain list; S3: Based on the split task chain list, extract the standby channel capacity in the transaction clearing channel, compare the remaining execution time of the task and the dependency weight, and select the pushable task according to the capacity in turn and assign it to the target channel, and obtain the standby channel push list; S4: Based on the standby channel push list, according to the task business type, currency type and compliance time limit, associate the currency type and time limit to the business type, match the regional supervision priority table, assign the processing order, and obtain the cross-regional priority processing list; S5: Based on the cross-regional priority processing list, read the number of branch nodes, compare the branch list and delay cumulative value, filter the unarrived branch and insert the position according to the delay value, and obtain the confluence node execution path scheme; The task dependency chain data refers to reading the execution time of each task and the task identifier based on the node task execution sequence, comparing in order, extracting the preposition between tasks, and arranging the dependency relationship in the original order; The dependency weight refers to reading the weight value of the task in the dependency chain when comparing the standby channel capacity information and the remaining execution time of the task, and the weight value is derived from the original data record of the task dependency relationship, which can be used as a task scheduling attribute in subsequent steps to determine the priority in combination with the standby channel capacity and the remaining execution time.

2. The method of claim 1, wherein, The node task execution sequence includes task entering sequence, task node position, task waiting time and processing delay comparison result, the split task chain list includes splittable task position, task dependency link information, sequence order comparison result, the standby channel push list includes task remaining execution time sorting result, available standby channel capacity allocation result and push task list, the cross-regional priority processing list includes currency type associated with business type, compliance time limit corresponding to currency type and regional supervision priority, and the confluence node execution path scheme includes unarrived branch list, delay cumulative value corresponding insertion position and necessary branch comparison result.

3. The method of claim 1, wherein, The supervision priority table refers to after the business type compliance association table is established, the regional code in the business type matching supervision priority table is matched, the matched regional information is combined to the task record, and the grouping and sorting are performed according to the region; The branch list refers to after the cross-regional priority processing list is established, the necessary branch identifier sequence of the confluence node is read, the completed branch and the necessary branch are compared, the unarrived branch is filtered, and the insertion position is inserted according to the delay cumulative value.

4. The method of claim 1, wherein, The specific steps of S1 are: S101: Obtain real-time processing delay, current task queuing quantity and node capacity data of a transaction node in a unified marketing scheduling platform, partition the data according to node identification, correspond the processing delay and the task queuing quantity in the partition to the node capacity, and obtain node partition performance information; S102: Based on the node partition performance information, according to the task queuing quantity and the node capacity of each node, and in contrast with the processing delay of the node, distribute the contrast result in the order of waiting time length, and obtain a node waiting time length sequence; S103: Based on the node waiting time length sequence, match each task to the corresponding node position in the order of entering, and connect the execution process according to the position of the node in the sequence, and obtain a node task execution sequence.

5. The method of claim 1, wherein, The specific steps of S2 are: S201: Based on the node task execution sequence, read the task execution dependency chain data, compare the task execution time and the task identification in sequence, verify the position of each task in the sequence, and obtain task order corresponding information; S202: Based on the task order corresponding information, filter the task positions that can be split in the order, calculate the task position node occupancy rate, and list the task positions in a list, and obtain a list of splittable task positions; S203: Based on the list of splittable task positions, extract link information consistent with the list position in the task execution dependency chain data, and connect according to the relationship between tasks in the order, and obtain a split task chain list.

6. The method of claim 1, wherein, The specific steps of S3 are: S301: Based on the split task chain list, extract the available capacity of the standby channel buffer in the financial transaction clearing channel, partition the capacity data according to the channel identification, and correspond the capacity and the task position, and obtain standby channel capacity information; S302: Based on the standby channel capacity information, compare the remaining execution time of each task with the task dependency weight, calculate the task dependency weight occupancy ratio, and list the order according to the remaining execution time, and obtain a task execution order table; S303: Based on the task execution order table, select the tasks that can be pushed to the standby channel in turn, assign the tasks to the corresponding target channel, and obtain a standby channel push list.

7. The method of claim 1, wherein, The specific steps of S4 are: S401: Based on the standby channel push list, read the transaction task business type, transaction currency type and compliance time limit, correspond the transaction currency type and compliance time limit according to the business type, and associate the relationship according to the task identification, and obtain a business type compliance association table; S402: Based on the business type compliance association table, match the business type to the corresponding area in the regulatory priority table, combine the area code matched to the original associated information, and obtain an area matching result table; S403: Based on the area matching result table, arrange the task processing order according to the compliance time limit, and assign the tasks to the corresponding area, and obtain a cross-area priority processing list.

8. The method of claim 1, wherein, The specific steps of S5 are: S501: Based on the cross-area priority processing list, read the completed branch quantity parameter of the merging node, correspond the quantity to the merging node required branch list parameter, identify the branch arrival state, and obtain a branch arrival state table; S502: Based on the branch arrival state table, the unarrived branch is compared with the delay cumulative value parameter to distribute the branch in the list according to the delay cumulative value, and an unarrived branch delay position table is obtained; S503: Based on the unarrived branch delay position table, the branch is inserted into the corresponding position in the confluence node required branch list parameter, and is connected into an executable node path to obtain a confluence node execution path scheme.

9. A digital one-net financial marketing management system, characterized in that, The system is used to realize the digital one-network financial marketing management method of any one of claims 1-8, and the system comprises: The transaction node monitoring module obtains the real-time processing delay, task queuing quantity and processing capacity of the transaction node in the unified marketing scheduling platform, compares the queuing quantity and processing capacity according to the node grouping, compares the waiting time and processing delay, and distributes the task entering and leaving according to the comparison result and the node position to obtain the node task execution sequence; The task link decomposition module reads the task dependency chain data and execution time based on the node task execution sequence, compares the execution time and order, identifies the separable task position, and extracts the corresponding link to obtain a split task chain list; The standby channel allocation module extracts the standby channel capacity of the financial transaction clearing channel based on the split task chain list, compares the remaining execution time and dependency weight of the task, and selects the pushable task according to the capacity to be distributed to the target channel in turn to obtain a standby channel push list; The cross-region priority processing module associates the currency type and time limit to the business type according to the transaction task business type, currency type and compliance time limit, matches the region according to the supervision priority table, and distributes the processing order to obtain a cross-region priority processing list; The confluence path planning module calls the completed branch quantity, required branch list and delay cumulative value of the confluence node based on the cross-region priority processing list, compares the branch list and delay cumulative value, filters the unarrived branch and inserts the position according to the delay value to obtain a confluence node execution path scheme.

Images