Account balance early warning method, device and equipment

Abstract

The application discloses a kind of account balance early warning method, device and equipment, the method is applied to logistics technical field, the method includes: the historical external call operation and maintenance data of target express messenger is collected, and multi-dimensional time series dataset is constructed;After processing multi-dimensional time series dataset, obtain the processed dataset;The processed dataset is input into time sequence attention TCN-LSTM hybrid prediction model, and the short-time mutation feature and long-time sequence periodic characteristic in the processed dataset are extracted;Short-time mutation feature and long-time sequence periodic characteristic are weighted and fused, and the balance consumption trend of target express messenger in future preset length is predicted;And determine the remaining balance in future preset length;If remaining balance is less than or equal to preset shutdown threshold, early warning information is pushed to the terminal of target express messenger.The application issues early warning information when the account balance of express messenger is insufficient, avoids sudden balance shortage shutdown, causes the problem of express delivery delay.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to an account balance early warning method, apparatus, and device. Background Technology

[0002] The following problems exist in the current technology for couriers using outbound calling services: 1. Passive management, lack of predictive ability, and frequent service interruptions. Existing systems can only monitor the current account balance in real time and cannot predict future balance consumption trends. Couriers cannot know in advance of the risk of insufficient balance. Outbound calling services are characterized by concentrated periods and large volumes. During peak periods, a large number of outbound call charges will be generated in a short period of time, which can easily lead to sudden insufficient balance and shutdown, causing forced interruption of ongoing delivery notifications, emergency package reminders, and other services, resulting in problems such as package delivery delays and customer complaints. 2. No automated recharge mechanism, resulting in high manual maintenance costs. The existing solution relies entirely on couriers to manually check the balance and recharge manually. Couriers do not have time to monitor the account status in real time during field operations. Recharging and restarting the system after a shutdown will create a long service gap, seriously affecting the continuity of outbound calling services. 3. Single and non-redundant resources, resulting in poor scheduling flexibility. The existing system uses a single salesperson account for independent operation, lacking the concept of a shared resource pool for branches. When a single salesperson's balance is insufficient, all outbound call tasks under their name are suspended, making it impossible to use idle resources as a backup. This results in extremely low resource utilization and poor system resilience. 4. The lack of task tiering and risk control mechanisms leads to an imbalance between efficiency and security. The existing technology does not differentiate between tasks of different priorities, such as urgent fresh produce deliveries, complaint warnings, and ordinary delivery orders. When the balance is insufficient, both high-priority and low-priority tasks are interrupted simultaneously. Furthermore, it lacks the ability to identify abnormal outbound calls, failing to intercept frequent repeated calls from the same number and other violations, which can easily lead to user complaints and operator risk control bans. Summary of the Invention

[0003] This invention provides an account balance early warning method, device, and equipment that can issue early warning information when a courier's account balance is insufficient, thereby avoiding sudden shutdowns due to insufficient balance and resulting in delivery delays.

[0004] On one hand, the present invention provides an account balance early warning method, the method comprising: Collect historical outbound call operation and maintenance data of the target courier to construct a multi-dimensional time-series dataset; after normalization, noise reduction and time-series alignment preprocessing of the multi-dimensional time-series dataset, the processed dataset is obtained. The processed dataset is input into the temporal attention TCN-LSTM hybrid prediction model, which includes a TCN module and an LSTM module. The TCN module extracts short-term mutation features from the processed dataset, and the LSTM module extracts long-term periodic features from the processed dataset. Based on the feature attention mechanism and the temporal attention mechanism, the short-term mutation features and the long-term periodic features are weighted and fused to predict the balance consumption trend of the target courier within a preset time period in the future. Based on the predicted balance consumption trend, the remaining balance within the preset future time period is determined; if the remaining balance is less than or equal to the preset shutdown threshold, a warning message is pushed to the target courier's terminal.

[0005] On the other hand, an account balance early warning device is provided, the device comprising: The data acquisition module is used to collect historical outbound call operation and maintenance data of the target courier and construct a multi-dimensional time series dataset. After normalization, noise reduction and time series alignment preprocessing of the multi-dimensional time series dataset, the processed dataset is obtained. The data input module is used to input the processed dataset into the temporal attention TCN-LSTM hybrid prediction model, which includes a TCN module and an LSTM module. The feature extraction module is used to extract short-term mutation features from the processed dataset through the TCN module and to extract long-term periodic features from the processed dataset through the LSTM module. The feature fusion module is used to perform weighted fusion of the short-term abrupt features and the long-term periodic features based on the feature attention mechanism and the temporal attention mechanism to predict the balance consumption trend of the target courier within a preset time period in the future. The early warning module is used to determine the remaining balance within the preset time period based on the prediction result of the balance consumption trend; if the remaining balance is less than or equal to the preset shutdown threshold, an early warning message is pushed to the terminal of the target courier.

[0006] In one exemplary embodiment, the early warning module is further configured to: if the remaining balance is detected to be less than or equal to a preset shutdown threshold, generate an insufficient balance message, an estimated shutdown time, and a manual recharge prompt; Based on the insufficient balance message, estimated downtime, and manual top-up prompt, a pop-up warning and reminder message are generated; and the pop-up warning and reminder message are pushed to the target courier's terminal.

[0007] In one exemplary embodiment, the apparatus further includes: The push module is used to perform a preset value recharge operation on the target courier's account if no account recharge operation is detected within a preset response window period after the warning information is pushed; the preset value is less than the target value. The deduction module is used to obtain the next day's settlement salary of the target courier and deduct the preset amount from the next day's settlement salary.

[0008] In one exemplary embodiment, the feature extraction module is further configured to extract short-term peak-period batch outbound call features and local sudden charge features from the processed dataset through the TCN module to obtain short-term mutation features; The LSTM module extracts the periodic features of balance consumption for different time periods in the processed dataset to obtain long-term periodic features.

[0009] In one exemplary embodiment, the apparatus further includes: The priority acquisition module is used to acquire the priority of the outbound call task when the remaining balance is detected to be less than a preset threshold or about to drop to the preset shutdown threshold. The task execution module is used to execute the outbound call task based on the target courier's account if the priority is high priority; The task migration module is used to migrate the outbound call task to the branch shared AI proxy call account for execution if the priority is low. The task rollback module is used to automatically roll back the outbound call task after detecting that the target courier's account has returned to normal.

[0010] In one exemplary embodiment, the high-priority outbound call tasks include tasks corresponding to fresh cold chain shipments, overdue warnings, and user complaint follow-up shipments; the low-priority outbound call tasks include ordinary delivery notifications and routine reminders.

[0011] In one exemplary embodiment, the apparatus further includes: The data monitoring module is used to monitor the behavior data of all outbound calling tasks in real time. If abnormal behavior is detected, such as repeated dialing of the same phone number in a short period of time, the difference between the outbound calling frequency in a single period and the historical average is greater than a preset value, or empty dialing, the outbound calling task is identified as an abnormal outbound calling task. The work order push module is used to automatically terminate the abnormal outbound call task, freeze the scheduling permission of the abnormal outbound call task, and push the abnormal work order corresponding to the abnormal outbound call task to the branch administrator's backend for manual review and investigation.

[0012] On the other hand, an electronic device is provided, the device including a processor and a memory, the memory storing at least one instruction or at least one program, the at least one instruction or the at least one program being loaded and executed by the processor to implement the account balance warning method as described above.

[0013] On the other hand, a computer storage medium is provided that stores at least one instruction or at least one program, which is loaded and executed by a processor to implement the account balance warning method as described above.

[0014] On the other hand, a computer program product or computer program is provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the account balance warning method as described above.

[0015] The account balance early warning method, apparatus, and device provided by this invention have the following technical effects: This invention collects historical outbound call operation data of target couriers to construct a multi-dimensional time-series dataset. After normalization, denoising, and time-series alignment preprocessing, a processed dataset is obtained. This processed dataset is then input into a temporal attention TCN-LSTM hybrid prediction model, which includes both a TCN module and an LSTM module. The TCN module extracts short-term mutation features from the processed dataset, and the LSTM module extracts long-term periodic features. Based on feature attention and temporal attention mechanisms, the short-term mutation features and long-term periodic features are weighted and fused to predict the target courier's balance consumption trend within a preset future timeframe. Then, based on the predicted balance consumption trend, the remaining balance within the preset future timeframe is determined. If the remaining balance is less than or equal to a preset shutdown threshold, a warning message is pushed to the target courier's terminal. This allows for early warnings when the courier's account balance is insufficient, preventing sudden balance shortages that could force interruptions of ongoing delivery notifications, emergency alerts, and other services, leading to delivery delays and customer complaints. Attached Figure Description

[0016] To more clearly illustrate the technical solutions and advantages in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the 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.

[0017] Figure 1 This is a schematic diagram of an account balance early warning system provided in the embodiments of this specification; Figure 2 This is a flowchart illustrating an account balance early warning method provided in the embodiments of this specification; Figure 3 This is a flowchart illustrating a task execution method provided in an embodiment of this specification; Figure 4 This is a flowchart illustrating a method for pushing early warning information to the terminal of the target courier, as provided in an embodiment of this specification. Figure 5 This is a flowchart illustrating a method for recharging the account of the target courier with a preset value, as provided in an embodiment of this specification. Figure 6 This is a schematic diagram of the structure of an account balance early warning device provided in the embodiments of this specification; Figure 7 This is a schematic diagram of the structure of a server provided in an embodiment of this specification. Detailed Implementation

[0018] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0020] Please see Figure 1 , Figure 1 This is a schematic diagram of an account balance early warning system provided in the embodiments of this specification, such as... Figure 1 As shown, the account balance early warning system may include at least server 01 and client 02.

[0021] Specifically, in the embodiments of this specification, server 01 may include a standalone server, a distributed server, or a server cluster composed of multiple servers. It may also be a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. Server 01 may include a network communication unit, a processor, and a memory, etc. Specifically, server 01 can be used to predict the balance consumption trend of the target courier within a preset future period based on the target courier's historical outbound call operation data; determine the remaining balance within the preset future period based on the predicted balance consumption trend; and send a push warning message to the target courier's client 02 if the remaining balance is less than or equal to a preset shutdown threshold.

[0022] Specifically, in the embodiments of this specification, the client 02 may include physical devices such as smartphones, desktop computers, tablets, laptops, digital assistants, smart wearable devices, smart speakers, in-vehicle terminals, and smart TVs. It may also include software running on the physical device, such as web pages provided to users by service providers, or applications provided to users by such service providers. Specifically, the client 02 can be used to receive and display push notification alerts.

[0023] The following describes an account balance early warning method according to the present invention. Figure 2 This is a flowchart illustrating an account balance early warning method provided in an embodiment of this specification. This specification provides the operational steps of the method described in the embodiment or flowchart, but based on conventional or non-inventive labor, more or fewer operational steps may be included. The order of steps listed in the embodiment is merely one possible execution order among many and does not represent the only possible execution order. In actual system or server product execution, the method can be executed sequentially or in parallel (e.g., in a parallel processor or multi-threaded processing environment) as shown in the embodiment or accompanying drawings. Specifically, as shown... Figure 2 As shown, the method may include: S201: Collect historical outbound call operation and maintenance data of the target courier and construct a multi-dimensional time series dataset; after normalizing, denoising and time-series alignment preprocessing of the multi-dimensional time series dataset, the processed dataset is obtained. In the embodiments of this specification, historical outbound call operation and maintenance data may include historical outbound call frequency, operating period, call connection rate, single call duration, average daily outbound call volume, average time period charge, and non-connection call charge data.

[0024] S203: Input the processed dataset into the temporal attention TCN-LSTM hybrid prediction model, which includes a TCN module and an LSTM module.

[0025] In the embodiments described in this specification, the TCN-LSTM hybrid prediction model includes a TCN module and an LSTM module.

[0026] S205: Extract short-term mutation features from the processed dataset using the TCN module, and extract long-term periodic features from the processed dataset using the LSTM module.

[0027] In the embodiments of this specification, the model completes local fine-grained feature extraction through the TCN temporal convolution module, accurately capturing sudden features such as batch outbound calls and sudden charges during short-term peak periods; and mines long-term periodic variation patterns through the LSTM module, capturing the periodic features of balance consumption brought about by daily and weekly operating habits.

[0028] S207: Based on the feature attention mechanism and the temporal attention mechanism, the short-term mutation features and the long-term periodic features are weighted and fused to predict the balance consumption trend of the target courier within a preset time period in the future.

[0029] In the embodiments of this specification, a feature attention mechanism and a temporal attention mechanism are added to adaptively strengthen the weight of high-impact factors (empty call charges, high-frequency outbound calls, low connection rate), suppress redundant and ineffective features by reducing their weight, iteratively optimize the model weight parameters, and accurately predict the account balance consumption trend, remaining balance value and insufficient balance risk period within a preset time period in the future.

[0030] S209: Based on the prediction results of the balance consumption trend, determine the remaining balance within the preset future time period; if the remaining balance is less than or equal to the preset shutdown threshold, push early warning information to the target courier's terminal.

[0031] In the embodiments described in this specification, based on the predicted balance consumption trend, when it is determined that the account balance will fall below a preset shutdown threshold within a preset future period, a terminal warning message is pushed. Both the preset future period and the preset shutdown threshold can be set according to actual circumstances.

[0032] In the embodiments of this specification, the temporal attention TCN-LSTM hybrid prediction model can adopt a sliding window temporal training method. For example, the training set can be the historical data of the past 30 days, the validation set can be the data of the past 3 days, and the preset prediction duration can be 2 hours. The model gradient propagation is optimized by bidirectional residual connection to avoid the gradient vanishing problem of deep models. It is suitable for the high-frequency operation and large time fluctuation of express delivery field operations. Compared with the traditional single LSTM model, it significantly improves the prediction accuracy of sudden change scenarios and ensures that the prediction results are consistent with the actual business scenario.

[0033] In some embodiments, the step of extracting short-term mutation features from the processed dataset using the TCN module and extracting long-term periodic features from the processed dataset using the LSTM module includes: The TCN module extracts short-term peak-period batch outbound call features and local sudden charge features from the processed dataset to obtain short-term mutation features. The LSTM module extracts the periodic features of balance consumption for different time periods in the processed dataset to obtain long-term periodic features.

[0034] In the embodiments of this specification, local fine-grained feature extraction can be completed through the TCN temporal convolution module to accurately capture sudden change features such as batch outbound calls and sudden charges during short-term peak periods, and obtain short-term sudden change features; the LSTM module can be used to mine long-term periodic change patterns to capture the periodic features of balance consumption brought about by daily and weekly operating habits, and obtain long-term periodic features.

[0035] In some embodiments, such as Figure 3 As shown, the method further includes: S301: If an outbound call task is detected, and the remaining balance is less than a preset threshold or about to drop to the preset shutdown threshold, the priority of the outbound call task is obtained. S303: If the priority is high priority, execute the outbound call task based on the target courier's account; S305: If the priority is low, the outbound call task will be migrated to the branch shared AI proxy call account for execution; S307: Automatically revert the outbound call task after detecting that the target courier's account has returned to normal.

[0036] In the embodiments described in this specification, the high priority and low priority can be set according to the actual situation.

[0037] This embodiment establishes a branch-level shared AI outbound calling resource pool. The pool aggregates idle public AI outbound calling accounts with sufficient balances within the branch, serving as a redundant backup resource for individual salesperson accounts. When a courier's account balance is detected to be insufficient or about to trigger the shutdown threshold, high-priority outbound calling tasks are retained for execution in the original account, while low-priority outbound calling tasks are automatically migrated to the branch's shared AI proxy calling account for execution. Once the original account returns to normal, the tasks are automatically migrated back.

[0038] In some embodiments, the high-priority outbound call tasks include tasks corresponding to fresh cold chain shipments, overdue warnings, and user complaint follow-up shipments; the low-priority outbound call tasks include ordinary delivery notifications and routine reminders.

[0039] (1) High-priority task retention: Fresh cold chain shipments, overdue warning shipments, and user complaint follow-up shipments are defined as high-priority tasks. These tasks are always retained in the original salesperson's account to ensure that emergency business is not interrupted. (2) Low-priority task migration: Low-priority tasks such as ordinary delivery notices and routine reminders will be automatically migrated to the branch's shared AI proxy account for execution. Once the original salesperson's account is recharged and restored to normal, the task will be automatically switched back to the original account. (3) Dynamic resource recycling: When the shared account task load reaches the preset threshold or the original account returns to normal, the system dynamically recycles the scheduled tasks to balance the outbound call resource load of the entire network.

[0040] In some embodiments, such as Figure 4 As shown, if the remaining balance is less than or equal to a preset shutdown threshold, pushing a warning message to the target courier's terminal includes: S401: If the remaining balance is detected to be less than or equal to a preset shutdown threshold, generate an insufficient balance message, estimated shutdown time, and manual recharge prompt; S403: Based on the insufficient balance message, estimated downtime, and manual recharge prompt, generate a pop-up warning and reminder message; and push the pop-up warning and reminder message to the target courier's terminal.

[0041] The preset shutdown threshold can be set to 2 yuan, the preset duration can be set to 2 days, and can be adjusted according to the actual situation.

[0042] In some embodiments, such as Figure 5 As shown, the method further includes: S501: If no account top-up operation is detected for the target courier within the preset response window period after the warning information is pushed, a preset value top-up operation is performed on the target courier's account; the preset value is less than the target value. S503: Obtain the next day's settlement salary of the target courier, and deduct the preset amount from the next day's settlement salary.

[0043] In the embodiments described in this specification, the target value can be set to 10 yuan. The recharge amount is included in the employee's daily maintenance bill and automatically deducted from the next day's salary settlement without manual intervention, achieving seamless recharge for the business.

[0044] In this embodiment of the specification, two levels of risk assessment thresholds can be set to implement tiered early warning and recharge strategies: (1) Warning Trigger: When the model predicts that the balance of the courier's personal account will fall below the 2 yuan suspension threshold within the next 2 hours, the system will immediately push a pop-up warning and a message push to the courier's terminal APP to inform him of the balance risk, the expected suspension time, and prompt him to recharge manually; (2) Automatic recharge trigger: After the warning is pushed, a 10-minute response window is preset. If no manual recharge operation is detected within the window, the system will automatically perform a small-amount, seamless recharge operation. The default recharge amount is 10 yuan. The recharge amount is included in the employee's daily maintenance bill and will be automatically deducted from the next day's salary settlement without manual intervention, thus achieving seamless recharge for the business.

[0045] In some embodiments, the method further includes: Real-time monitoring of the behavior data of all outbound calling tasks. If abnormal behavior is detected, such as repeated dialing of the same phone number in a short period of time, the difference between the outbound calling frequency in a single period of time and the historical average is greater than a preset value, or empty dialing, the outbound calling task is identified as an abnormal outbound calling task. The abnormal outbound call task is automatically terminated, the scheduling permission for the abnormal outbound call task is frozen, and the abnormal work order corresponding to the abnormal outbound call task is pushed to the branch administrator's backend for manual review and investigation.

[0046] In the embodiments described in this specification, all AI outbound call behavior data are monitored in real time to identify abnormal high-frequency outbound call behavior. When abnormal rules are triggered, outbound call tasks are automatically suspended, scheduling permissions are locked, and manual review is requested.

[0047] The system monitors all outbound call data in real time and identifies abnormal high-frequency outbound call scenarios. If abnormal behaviors such as repeatedly dialing the same phone number in a short period of time, outbound call frequency far exceeding the historical average in a single time period, or meaningless empty dialing are detected, the abnormal circuit breaker mechanism is immediately triggered to automatically terminate the abnormal outbound call task, freeze the corresponding scheduling permissions, and push the abnormal work order to the branch administrator's backend for manual review and investigation, thereby avoiding the risks of operator risk control blocking and user complaints.

[0048] As can be seen from the technical solutions provided in the embodiments of this specification above, the embodiments of this specification collect historical outbound call operation and maintenance data of target couriers to construct a multi-dimensional time-series dataset; after normalization, denoising, and time-series alignment preprocessing of the multi-dimensional time-series dataset, a processed dataset is obtained; the processed dataset is input into a time-series attention TCN-LSTM hybrid prediction model, the hybrid prediction model including a TCN module and an LSTM module; short-term mutation features are extracted from the processed dataset through the TCN module, and long-term periodic features are extracted from the processed dataset through the LSTM module; based on feature attention... The mechanism and the temporal attention mechanism perform weighted fusion of the short-term mutation features and the long-term periodic features to predict the balance consumption trend of the target courier within a preset time period in the future; then, based on the prediction result of the balance consumption trend, the remaining balance within the preset time period in the future is determined; if the remaining balance is less than or equal to a preset shutdown threshold, an early warning information is pushed to the target courier's terminal, so that an early warning information can be issued in advance when the courier's account balance is insufficient, avoiding sudden insufficient balance shutdown, which would cause the forced interruption of ongoing delivery notifications, emergency item reminders and other services, resulting in problems such as delayed delivery and user complaints.

[0049] This specification also provides an account balance early warning device, such as... Figure 6 As shown, the device includes: The data acquisition module 610 is used to collect historical outbound call operation and maintenance data of the target courier and construct a multi-dimensional time series dataset; after normalization, noise reduction and time series alignment preprocessing of the multi-dimensional time series dataset, the processed dataset is obtained. Data input module 620 is used to input the processed dataset into the temporal attention TCN-LSTM hybrid prediction model, the hybrid prediction model including a TCN module and an LSTM module; The feature extraction module 630 is used to extract short-term mutation features from the processed dataset through the TCN module and to extract long-term periodic features from the processed dataset through the LSTM module. The feature fusion module 640 is used to perform weighted fusion of the short-term mutation features and the long-term periodic features based on the feature attention mechanism and the temporal attention mechanism to predict the balance consumption trend of the target courier within a preset time period in the future. The early warning module 650 is used to determine the remaining balance within the future preset time period based on the prediction result of the balance consumption trend; if the remaining balance is less than or equal to a preset shutdown threshold, it pushes early warning information to the terminal of the target courier.

[0050] In one exemplary embodiment, the early warning module is further configured to: if the remaining balance is detected to be less than or equal to a preset shutdown threshold, generate an insufficient balance message, an estimated shutdown time, and a manual recharge prompt; Based on the insufficient balance message, estimated downtime, and manual top-up prompt, a pop-up warning and reminder message are generated; and the pop-up warning and reminder message are pushed to the target courier's terminal.

[0051] In one exemplary embodiment, the apparatus further includes: The push module is used to perform a preset value recharge operation on the target courier's account if no account recharge operation is detected within a preset response window period after the warning information is pushed; the preset value is less than the target value. The deduction module is used to obtain the next day's settlement salary of the target courier and deduct the preset amount from the next day's settlement salary.

[0052] In one exemplary embodiment, the feature extraction module is further configured to extract short-term peak-period batch outbound call features and local sudden charge features from the processed dataset through the TCN module to obtain short-term mutation features; The LSTM module extracts the periodic features of balance consumption for different time periods in the processed dataset to obtain long-term periodic features.

[0053] In one exemplary embodiment, the apparatus further includes: The priority acquisition module is used to acquire the priority of the outbound call task when the remaining balance is detected to be less than a preset threshold or about to drop to the preset shutdown threshold. The task execution module is used to execute the outbound call task based on the target courier's account if the priority is high priority; The task migration module is used to migrate the outbound call task to the branch shared AI proxy call account for execution if the priority is low. The task rollback module is used to automatically roll back the outbound call task after detecting that the target courier's account has returned to normal.

[0054] In one exemplary embodiment, the high-priority outbound call tasks include tasks corresponding to fresh cold chain shipments, overdue warnings, and user complaint follow-up shipments; the low-priority outbound call tasks include ordinary delivery notifications and routine reminders.

[0055] In one exemplary embodiment, the apparatus further includes: The data monitoring module is used to monitor the behavior data of all outbound calling tasks in real time. If abnormal behavior is detected, such as repeated dialing of the same phone number in a short period of time, the difference between the outbound calling frequency in a single period and the historical average is greater than a preset value, or empty dialing, the outbound calling task is identified as an abnormal outbound calling task. The work order push module is used to automatically terminate the abnormal outbound call task, freeze the scheduling permission of the abnormal outbound call task, and push the abnormal work order corresponding to the abnormal outbound call task to the branch administrator's backend for manual review and investigation.

[0056] The apparatus and method embodiments described herein are based on the same inventive concept.

[0057] This specification provides an electronic device including a processor and a memory. The memory stores at least one instruction or at least one program, which is loaded and executed by the processor to implement the account balance warning method provided in the above method embodiments.

[0058] Embodiments of the present invention also provide a computer storage medium, which can be disposed in a terminal to store at least one instruction or at least one program related to implementing an account balance warning method in the method embodiments. The at least one instruction or at least one program is loaded and executed by the processor to implement the account balance warning method provided in the above method embodiments.

[0059] Embodiments of the present invention also provide a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the account balance warning method provided in the above-described method embodiments.

[0060] Optionally, in the embodiments of this specification, the storage medium may be located at at least one of the multiple network servers in a computer network. Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0061] The memory described in the embodiments of this specification can be used to store software programs and modules. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for the functions, etc.; the data storage area may store data created according to the use of the device, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory may also include a memory controller to provide the processor with access to the memory.

[0062] The account balance early warning method embodiments provided in this specification can be executed on mobile terminals, computer terminals, servers, or similar computing devices. Taking running on a server as an example, Figure 7 This is a hardware structure block diagram of a server for an account balance early warning method provided in the embodiments of this specification. For example... Figure 7 As shown, the server 700 can vary significantly due to different configurations or performance. It may include one or more Central Processing Units (CPUs) 710 (CPUs 710 may include, but are not limited to, microprocessors (MCUs) or programmable logic devices (FPGAs), a memory 730 for storing data, and one or more storage media 720 (e.g., one or more mass storage devices) for storing application programs 723 or data 722. The memory 730 and storage media 720 may be temporary or persistent storage. The program stored in the storage media 720 may include one or more modules, each module may include a series of instruction operations on the server. Furthermore, the CPU 710 may be configured to communicate with the storage media 720 and execute the series of instruction operations stored in the storage media 720 on the server 700. Server 700 may also include one or more power supplies 760, one or more wired or wireless network interfaces 750, one or more input / output interfaces 740, and / or one or more operating systems 721, such as Windows Server™, Mac OS X™, Unix™, Linux™, FreeBSD™, etc.

[0063] The input / output interface 740 can be used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of server 700. In one example, the input / output interface 740 includes a network interface controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the input / output interface 740 may be a radio frequency (RF) module used for wireless communication with the Internet.

[0064] Those skilled in the art will understand that Figure 7 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, server 700 may also include... Figure 7 The more or fewer components shown, or having the same Figure 7 The different configurations shown.

[0065] As can be seen from the embodiments of the account balance early warning method, device, electronic device or storage medium provided by the present invention, the present invention collects historical outbound call operation and maintenance data of the target courier to construct a multi-dimensional time series dataset; after normalizing, denoising and time-series alignment preprocessing of the multi-dimensional time series dataset, a processed dataset is obtained; the processed dataset is input into a time-series attention TCN-LSTM hybrid prediction model, the hybrid prediction model including a TCN module and an LSTM module; the TCN module extracts short-term mutation features in the processed dataset, and the LSTM module extracts long-term periodic features in the processed dataset; The short-term mutation features and the long-term periodic features are weighted and fused based on feature attention mechanism and temporal attention mechanism to predict the balance consumption trend of the target courier within a preset time period. Then, based on the prediction result of the balance consumption trend, the remaining balance within the preset time period is determined. If the remaining balance is less than or equal to a preset shutdown threshold, an early warning message is pushed to the target courier's terminal. This allows for early warning when the courier's account balance is insufficient, avoiding sudden shutdown due to insufficient balance, which could force the interruption of ongoing services such as delivery notifications and emergency reminders, causing problems such as delayed delivery and user complaints.

[0066] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, specific embodiments of this specification have been described above. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0067] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of apparatus, devices, and storage media are basically similar to the method embodiments, so the descriptions are relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0068] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware or by a program instructing related hardware. The program can be stored in a computer storage medium, such as a read-only memory, a disk, or an optical disk.

[0069] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for early warning of account balance, characterized in that, The method includes: Collect historical outbound call operation and maintenance data of the target courier to construct a multi-dimensional time-series dataset; after normalization, noise reduction and time-series alignment preprocessing of the multi-dimensional time-series dataset, the processed dataset is obtained. The processed dataset is input into the temporal attention TCN-LSTM hybrid prediction model, which includes a TCN module and an LSTM module. The TCN module extracts short-term mutation features from the processed dataset, and the LSTM module extracts long-term periodic features from the processed dataset. Based on the feature attention mechanism and the temporal attention mechanism, the short-term mutation features and the long-term periodic features are weighted and fused to predict the balance consumption trend of the target courier within a preset time period in the future. Based on the predicted balance consumption trend, the remaining balance within the preset future time period is determined; if the remaining balance is less than or equal to the preset shutdown threshold, a warning message is pushed to the target courier's terminal.

2. The method according to claim 1, characterized in that, If the remaining balance is less than or equal to a preset shutdown threshold, a warning message is pushed to the target courier's terminal, including: If the remaining balance is detected to be less than or equal to a preset shutdown threshold, an insufficient balance message, estimated shutdown time, and manual recharge prompt will be generated. Based on the insufficient balance message, estimated downtime, and manual top-up prompt, a pop-up warning and reminder message are generated; and the pop-up warning and reminder message are pushed to the target courier's terminal.

3. The method according to claim 1, characterized in that, The method further includes: If no account top-up operation is detected within a preset response window period after the warning information is pushed, a preset value top-up operation is performed on the target courier's account; the preset value is less than the target value. Obtain the next day's settlement salary for the target courier, and deduct the preset amount from the next day's settlement salary.

4. The method according to claim 1, characterized in that, The step of extracting short-term mutation features from the processed dataset using the TCN module and extracting long-term periodic features from the processed dataset using the LSTM module includes: The TCN module extracts short-term peak-period batch outbound call features and local sudden charge features from the processed dataset to obtain short-term mutation features. The LSTM module extracts the periodic features of balance consumption for different time periods in the processed dataset to obtain long-term periodic features.

5. The method according to claim 2, characterized in that, The method further includes: If an outbound call task is detected, and the remaining balance is less than a preset threshold or about to drop to the preset shutdown threshold, the priority of the outbound call task is obtained. If the priority is high, the outbound call task is executed based on the target courier's account; If the priority is low, the outbound call task will be migrated to the branch's shared AI proxy call account for execution; The outbound call task will be automatically reassigned after the target courier's account is detected to be back to normal.

6. The method according to claim 5, characterized in that, The high-priority outbound call tasks include tasks corresponding to fresh cold chain shipments, overdue warnings, and user complaint follow-ups; the low-priority outbound call tasks include ordinary delivery notifications and routine reminders.

7. The method according to any one of claims 1-6, characterized in that, The method further includes: Real-time monitoring of the behavior data of all outbound calling tasks. If abnormal behavior is detected, such as repeated dialing of the same phone number in a short period of time, the difference between the outbound calling frequency in a single period of time and the historical average is greater than a preset value, or empty dialing, the outbound calling task is identified as an abnormal outbound calling task. The abnormal outbound call task is automatically terminated, the scheduling permission for the abnormal outbound call task is frozen, and the abnormal work order corresponding to the abnormal outbound call task is pushed to the branch administrator's backend for manual review and investigation.

8. An account balance early warning device, characterized in that, The device includes: The data acquisition module is used to collect historical outbound call operation and maintenance data of the target courier and construct a multi-dimensional time series dataset. After normalization, noise reduction and time series alignment preprocessing of the multi-dimensional time series dataset, the processed dataset is obtained. The data input module is used to input the processed dataset into the temporal attention TCN-LSTM hybrid prediction model, which includes a TCN module and an LSTM module. The feature extraction module is used to extract short-term mutation features from the processed dataset through the TCN module and to extract long-term periodic features from the processed dataset through the LSTM module. The feature fusion module is used to perform weighted fusion of the short-term abrupt features and the long-term periodic features based on the feature attention mechanism and the temporal attention mechanism to predict the balance consumption trend of the target courier within a preset time period in the future. The early warning module is used to determine the remaining balance within the preset time period based on the prediction result of the balance consumption trend; if the remaining balance is less than or equal to the preset shutdown threshold, an early warning message is pushed to the terminal of the target courier.

9. An electronic device, characterized in that, The device includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, the at least one instruction or at least one program being loaded and executed by the processor to implement the account balance warning method as described in any one of claims 1-7.

10. A computer storage medium, characterized in that, The computer storage medium stores at least one instruction or at least one program, which is loaded and executed by a processor to implement the account balance early warning method as described in any one of claims 1-7.

Images