An event-driven settlement cycle automatic adjustment method and device

Abstract

The application discloses an event-driven settlement cycle automatic adjustment method and device, wherein the method comprises the following steps: receiving a plurality of risk events configured by a user through a visual interface; wherein the attributes of each event configured include a data source, a severity and a basic degradation gear number; detecting a risk event in real time through the data source, and generating a risk event record when detecting the risk event; determining a settlement cycle reduction time based on a preset formula, using the severity and the basic degradation gear number of each event in the risk event record; adjusting the settlement cycle of an enterprise using the settlement cycle reduction time; monitoring the risk event influence degree and the enterprise credit rating of the enterprise within a backtracking observation window, and issuing a warning or a ladder recovery settlement cycle according to the monitoring situation. The application can improve the settlement risk control ability and the airline user experience.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an event-driven method and apparatus for automatically adjusting the settlement cycle. Background Technology

[0002] In the air transport system, airports, as crucial infrastructure operators, provide airlines with a range of key services, including aircraft takeoff and landing, parking, passenger boarding, baggage handling, and ground service support. These services are typically not provided free of charge but are subject to corresponding fees based on service agreements signed by both parties. To standardize financial transactions and improve settlement efficiency, airports and airlines use settlement platforms to compile detailed information on all services provided by the airport and their corresponding fees within an agreed settlement period, generating a settlement invoice that is then deducted from the airline's prepayment account.

[0003] As the scale of airline settlement business expands, airlines' credit status and payment behavior are constantly changing. External credit rating agencies (such as IATA) will adjust airline ratings from time to time; airlines themselves may experience sudden financial deterioration (such as a sudden increase in debt, fluctuations in the home country's exchange rate), operational anomalies (such as large-scale flight cancellations), or public opinion risks (such as bankruptcy rumors). These risk events are often sudden, time-sensitive, and uncertain, requiring settlement platforms to respond quickly.

[0004] However, existing technologies primarily rely on manual configuration or periodic assessments (e.g., monthly) to adjust settlement cycles, failing to provide timely responses to the aforementioned unforeseen events. Furthermore, current systems can only identify existing fixed risk event categories for airlines, unable to directly monitor temporary or specific risk events (such as "overdue for more than 3 days" or "guarantor bank downgrade"), resulting in weak awareness of sudden risks. Additionally, existing solutions lack refined definitions and differentiated handling capabilities for risk event types, employing a "one-size-fits-all" approach to downgrade measures. This means that the same settlement cycle downgrade method is uniformly applied after an unforeseen event is detected, and adjustments to the settlement cycle require communication and coordination between both parties after the event's impact subsides. This leads to delayed risk control, high manual maintenance costs, and a poor airline experience. Summary of the Invention

[0005] This invention provides an event-driven automatic settlement cycle adjustment method to support configurable risk events, reassess settlement cycles based on different companies' risk events, and automatically restore settlement cycles, thereby reducing manual operation and maintenance costs, improving settlement risk control capabilities, and enhancing the airline user experience. The method includes: The system receives multiple risk events configured by the user through a visual interface; each event is configured with attributes including data source, severity, and base degradation level; the base degradation level reflects the range of reduction in the settlement cycle when the event is triggered individually. Risk events are detected in real time through the data source. When a risk event is detected, a risk event record is generated. The risk event record includes attribute information of a single event or a combination of events. Based on a preset formula, the reduction in settlement cycle duration is determined by utilizing the severity of each event in the risk event record and the base downgrade level. The preset formula is based on the base downgrade level, incorporates the severity of the event, and uses a mathematical function to quantify the reduction in settlement cycle duration. The reduced settlement cycle duration is used to adjust the company's settlement cycle. Within the retrospective observation window, monitor the impact of risk events on enterprises and their credit ratings, and issue early warnings or gradually restore settlement cycles based on the monitoring results.

[0006] This invention also provides an event-driven automatic settlement cycle adjustment device to support configurable risk events, reassess settlement cycles according to different enterprises' risk events, and achieve automatic recovery of settlement cycles, thereby reducing manual operation and maintenance costs, improving settlement risk control capabilities, and enhancing the airline user experience. The device includes: The risk event subscription configuration module is used to receive multiple risk events configured by the user through a visual interface; the attributes configured for each event include data source, severity, and base degradation level; the base degradation level reflects the reduction range of the settlement cycle when the event is triggered individually; The monitoring module is used to detect risk events in real time through the data source. When a risk event is detected, a risk event record is generated. The risk event record includes attribute information of a single event or a compound event. The instant reassessment module is used to determine the reduction in settlement cycle duration based on a preset formula, utilizing the severity of each event in the risk event record and the base downgrade level. The preset formula is based on the base downgrade level, incorporates the severity of the event, and uses a mathematical function to quantify the reduction in settlement cycle duration. The periodic degradation execution module is used to adjust the enterprise's settlement cycle by utilizing the reduced duration of the settlement cycle; The retrospective observation module is used to monitor the impact of risk events on enterprises and their credit rating within the retrospective observation window, and to issue early warnings or gradually restore the settlement cycle based on the monitoring results.

[0007] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described event-driven automatic adjustment method for settlement cycles.

[0008] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned event-driven automatic settlement cycle adjustment method.

[0009] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned event-driven automatic adjustment method for settlement cycles.

[0010] In this embodiment of the invention, multiple risk events configured by the user through a visual interface are received. Each event is configured with attributes including a data source, severity, and a base downgrade level. The base downgrade level reflects the reduction range of the settlement cycle when the event is triggered individually. Risk events are detected in real time through the data source, and a risk event record is generated when a risk event is detected. The risk event record includes attribute information of single or compound events. Based on a preset formula, the reduction duration of the settlement cycle is determined using the severity and base downgrade level of each event in the risk event record. The reduction duration of the settlement cycle is used to adjust the enterprise's settlement cycle. The impact of risk events on the enterprise and the enterprise's credit rating are monitored within a retrospective observation window, and warnings or tiered recovery of the settlement cycle are issued based on the monitoring results. In this embodiment of the invention, a visual interface allows users to configure risk events, enabling flexible configuration of event-driven risk control rules. Risk events are then detected in real time, and when detected, a risk event record is generated. The severity of each event in the record and the base downgrade level are used to determine the reduction in the settlement cycle duration. This allows for reassessment of settlement cycles based on the risk events of different companies. Furthermore, after the settlement cycle is adjusted, the impact of risk events on the company and its credit rating are monitored within a retrospective observation window. Based on the monitoring results, warnings are issued or the settlement cycle is restored in stages, enabling automatic recovery of the settlement cycle. This reduces manual maintenance costs, improves settlement risk control capabilities, and enhances the airline user experience. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the 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. In the drawings: Figure 1 This is a flowchart illustrating the event-driven automatic settlement cycle adjustment method in an embodiment of the present invention. Figure 2 This is a specific example diagram of the event-driven automatic adjustment method for settlement cycle in an embodiment of the present invention; Figure 3 This is another specific example of the event-driven automatic adjustment method for the settlement cycle in this invention. Figure 4 This is a schematic diagram of an event-driven automatic settlement cycle adjustment device according to an embodiment of the present invention; Figure 5 This is a schematic diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0012] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings. Here, the illustrative embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0013] The acquisition, storage, use, and processing of data in this application comply with relevant laws and regulations.

[0014] The existing settlement cycle adjustment scheme based on periodic assessment is as follows: Regular assessment mechanism: The system collects data such as the airline's historical payment timeliness rate, external credit rating, and current outstanding amount at fixed time intervals (e.g., the 1st of each month) to calculate a comprehensive risk score.

[0015] Threshold matching: The risk score is matched with a preset score range (e.g., excellent ≥85 points, normal 70-84 points, attention 50-69 points, high risk 30-49 points, severe risk <30 points), and each range corresponds to a fixed settlement period.

[0016] Periodic Update: If the current settlement period matched with the rating is different from the existing period, the system will update the airline's billing cycle configuration after the assessment date, and subsequent bills will be generated according to the new period.

[0017] Manual intervention: For individual sudden risk events (such as an airline being downgraded by IATA), the system does not handle them automatically and requires manual evaluation and modification of the cycle configuration by personnel.

[0018] Because existing technologies employ fixed-cycle evaluation and manual intervention, they suffer from the following drawbacks: Delayed response: After a risk event (such as an IATA rating downgrade or a sudden increase in arrears) occurs, the system must wait until the next scheduled assessment date (which may take several weeks) before the cycle can be adjusted. During this period, high-risk airlines continue to enjoy the original lenient cycle, and the risk exposure continues to expand, increasing the risk of bad debts.

[0019] Fixed event definition: The system can only identify indirect indicators such as score changes and cannot directly monitor specific risk events (such as "overdue for more than 3 days" or "downgrade of the guarantor bank"), resulting in a weak ability to perceive sudden risks.

[0020] Fixed downgrade range: Even if the rating triggers a cross-level downgrade, the downgrade range is uniquely determined by the rating range, and it is impossible to differentiate the treatment based on the severity of the event (e.g., minor delinquency and serious rating downgrades have the same downgrade range).

[0021] Unable to recover automatically: After the impact of the event subsides (such as the airline making up the debt or the rating being upgraded), the system still needs to wait for the next regular assessment date to reassess and communicate whether to resume the cycle. Moreover, the recovery process is not step-by-step and may easily lead to the cycle rebounding too quickly.

[0022] Lack of handling of complex events: When multiple risk events occur consecutively in a short period of time, existing technologies cannot make a combined judgment, which may underestimate the risk.

[0023] To address the aforementioned technical deficiencies, the embodiments of the present invention aim to solve the following technical problems caused by fixed-period evaluation and manual intervention in the prior art: 1. The response to serious risk events is delayed, and the risk exposure period is too long; 2. It is impossible to directly monitor and distinguish different risk events; the event definitions are fixed. 3. The downgrade level is applied uniformly and cannot be adjusted adaptively according to the severity of the event; 4. The settlement cycle cannot be automatically resumed after the impact of the event subsides, and manual operation is required; 5. When multiple events occur concurrently, there is a lack of a comprehensive judgment mechanism, which may lead to the omission of overlapping risks.

[0024] This invention provides an event-driven automatic adjustment method for settlement cycles, specifically an event-driven real-time reassessment and adaptive degradation method for settlement cycles, which enables real-time capture of risk events, refined hierarchical response, adaptive degradation and automated recovery, forming a closed-loop management of risk response.

[0025] Figure 1 This is a flowchart illustrating the event-driven automatic adjustment method for the settlement cycle in an embodiment of the present invention, as shown below. Figure 1 As shown, the method includes: Step 101: Receive multiple risk events configured by the user through a visual interface; the attributes configured for each event include data source, severity, and base downgrade level; the base downgrade level reflects the reduction range of the settlement cycle when the event is triggered individually; Step 102: Detect risk events in real time through the data source. When a risk event is detected, generate a risk event record. The risk event record includes attribute information of a single event or a compound event. Step 103: Based on a preset formula, determine the reduction duration of the settlement cycle by using the severity of each event in the risk event record and the base downgrade level; the preset formula is based on the base downgrade level, integrates the severity of the event, and uses a mathematical function to quantify the reduction duration of the settlement cycle. Step 104: Utilize the reduced settlement cycle duration to adjust the company's settlement cycle; Step 105: Monitor the impact of risk events on enterprises and their credit rating within the retrospective observation window, and issue early warnings or gradually restore settlement cycles based on the monitoring results.

[0026] The event-driven automatic adjustment method for settlement cycle in this embodiment of the invention will be explained in detail below.

[0027] In step 101, multiple risk events configured by the user through a visual interface are received.

[0028] In this embodiment of the invention, a configurable risk event subscription system is constructed.

[0029] An event subscription center is established, allowing users to dynamically configure various risk events to be monitored through a visual interface. As the central hub for rule management, the event subscription center aggregates the risk event types, monitoring scope, data sources, keywords, descriptions, attributes, and other conditions configured by users on the visual interface, ultimately generating risk rules.

[0030] Each event is configured with at least the following attributes: data source, severity, and base degradation level; the base degradation level reflects the range of reduction in the settlement cycle when the event is triggered individually.

[0031] For example, referring to Table 1, the events are categorized as follows: Table 1

[0032] Configure the following properties for each event: 1. Event ID, Event Name, Data Source, Trigger Threshold (Quantification Condition); 2. Severity coefficient S (0~1, representing the contribution weight of the event to the risk, e.g., severe = 1.0, moderate = 0.5, minor = 0.2). 3. Base downgrade level B, for example, the base downgrade level B represents the number of time intervals that should be shortened when the event is triggered alone, with a value range of 0 to 4 levels; each level corresponds to a different settlement period; 4. Effective status.

[0033] In the event configuration table 1 of step 101, each event requires a quantifiable trigger condition to determine whether the event has occurred. Refer to Table 2 below for examples of specific trigger thresholds for several different event categories: Table 2

[0034] A more detailed example – payment default events, assuming the system is configured as follows: Event ID: PAY_OVERDUE_3D; Event Name: Bill Overdue for More Than 3 Days; Data source: Settlement system - Accounts Receivable statement; Trigger threshold (quantification condition): { "Overdue days": ">= 3", "Unpaid bill status": "Overdue and unpaid", "Exclusion condition": "If the airline has submitted a dispute and the dispute status is 'processing', it will not trigger"}; Severity coefficient S: 0.8 (3 days overdue is relatively serious but not the most serious); Basic downgrade level B: 1 (at least 1 level downgrade for each overdue payment); When the system detects that an airline has a bill with an overdue date that has reached 3 days and there is no valid dispute, it determines that the event has been triggered and begins the subsequent immediate reassessment process.

[0035] More complex threshold examples (compound conditions): For the "sudden increase in outstanding payments" event, more granular thresholds can be set: Current total outstanding balance > 1.5 times the historical 30-day moving average outstanding balance AND absolute increase > 500,000; This avoids false triggers caused by small fluctuations, and only when the amount owed increases significantly and in absolute terms is it considered a risk event.

[0036] Therefore, in this embodiment of the invention, the trigger threshold is the quantified boundary condition for the system to determine "whether the risk event has occurred." It typically includes comparison operators (>, <, ≥, ≤, =), time windows, percentages, absolute values, and optional exclusion rules. With a clearly defined threshold, the system can automatically and unambiguously capture events.

[0037] In one embodiment, receiving multiple risk events configured by the user through a visual interface may include: configuring risk event categories, event subtypes, and event descriptions through drop-down selection, checkboxes, or custom input boxes; wherein the severity of the event is automatically recommended based on historical data statistics, rule template settings, or machine learning; the severity represents the weight of the event's contribution to the risk.

[0038] For example, users can monitor certain risk events for specific airlines through drop-down selections and checkboxes, and users can also customize new risk event categories, event subtypes, and event descriptions.

[0039] For example, for the severity coefficient S of some existing risk event categories, the initial value can be automatically output by statistically analyzing historical data, including the severity coefficient S of similar historical risk events. Alternatively, a rule mapping table (including the mapping relationship between risk events and severity coefficient S) can be predefined, or the severity coefficient S can be predicted by using a machine learning model based on risk event categories, keywords, and historical data.

[0040] The severity coefficient S is displayed on the interface via an input box or slider, and users can adjust it.

[0041] In step 102, event monitoring and real-time reassessment are triggered by: detecting risk events in real time through the data source; when a risk event is detected, a risk event record is generated; the risk event record includes attribute information of a single event or a compound event.

[0042] Connect to data sources and detect risk events in real time through message queues, stream computing engines, and other technologies.

[0043] Figure 2 This is a specific example diagram of the event-driven automatic adjustment method for the settlement cycle in an embodiment of the present invention, as shown below. Figure 2 As shown, risk events are detected in real time through the data source. When a risk event is detected, a risk event record is generated, which may include: Step 201: Obtain multi-source heterogeneous data in real time by polling each data source and / or subscribing to push data; Step 202: Perform entity recognition, semantic feature extraction, and risk rule matching on multi-source heterogeneous data through stream processing; the risk rules are obtained based on the configuration information configured by the user through the visual interface. Step 203: When a match is successful, a risk event record is generated; the risk event record includes a timestamp, enterprise identifier, risk event type, severity, and base downgrade level.

[0044] For example, for risk control platforms or regulatory interfaces that support push notifications, event-driven access is achieved using Webhook or Kafka subscriptions; for websites and news portals that do not have open push notification capabilities, text, images, and even short video content are periodically crawled using distributed polling crawlers. This raw data (including webpage text, OCR-recognized image text, ASR-transcribed video and audio, etc.) is uniformly sent into a streaming processing pipeline.

[0045] During the stream processing phase, the system invokes a multimodal large model to perform real-time parsing of heterogeneous content: For text, NER (Named Entity Recognition) and event extraction models are used to extract elements such as enterprise entity, behavior, time, and amount; For images, a large visual model is used in conjunction with OCR to recognize the text and scene semantics in the image and determine whether it contains risk signals such as announcements, penalty documents, and on-site photos. For video / audio, it is first transcribed into text using ASR, and then combined with image analysis for contextual understanding.

[0046] Subsequently, the extracted structured features are matched in real time with the risk rule base formed by user configuration. When a rule is matched, a risk event record is generated, including timestamp, airline ID, event type, severity coefficient S, basic downgrade level B, etc.

[0047] In one embodiment, risk events are detected in real time through the data source. When a risk event is detected, a risk event record is generated, including: Perform debouncing checks on risk events; the debouncing check means that if the same type of event from the same airline is triggered repeatedly within a preset time window (such as 1 hour), only the first one will be processed to avoid event storms.

[0048] When a risk event record is generated, an immediate reassessment process is triggered, without waiting for a periodic assessment cycle.

[0049] In step 103, the adaptive downgrade magnitude calculation is performed by determining the reduction duration of the settlement cycle based on a preset formula, using the severity of each event in the risk event record and the base downgrade level.

[0050] The preset formula is based on the number of basic downgrade levels, takes into account the severity of the event, and uses mathematical functions to quantify the reduction in the settlement cycle.

[0051] Mathematical functions, such as linear weighted models, can be used to calculate the settlement period as: Settlement period = Base duration corresponding to the base downgrade level B + k × Severity coefficient of the event, where k is an adjustment parameter.

[0052] Alternatively, an exponential / logarithmic decay model can be selected based on the sensitive scenario, but this requires multiple experimental verifications.

[0053] To improve computational efficiency and support rapid iteration and ease of maintenance, in one embodiment, the reduction in settlement cycle duration is determined based on a preset formula, utilizing the severity of each event in the risk event record and the base downgrade level. This may include: For a single event, Di is calculated using the following preset formula: D_i = ceil(B_i × S_i); where D_i represents the number of downgrade levels triggered independently by the i-th event, ceil() represents rounding up the result, B_i represents the severity coefficient, B_i∈[0,1], S_i represents the base downgrade level, S_i is any integer level between level 0 and level 4, where different levels correspond to different settlement periods of different time lengths; For composite events, D_total is calculated using the following preset formula: D_total = min(maximum downgrade limit, Σ(D_i × attenuation factor_k)); where D_total represents the number of downgrades, and the maximum downgrade limit represents the total number of downgrades from the current settlement period to the lowest settlement period. The attenuation factor _k = e^{-(k-1)×α}, where k is the sequence number of the event sorted by time (attenuation factor = 1 when k=1), and α is the configurable attenuation coefficient.

[0054] In the embodiments, D_i and D_total can be collectively referred to as the comprehensive degradation magnitude D.

[0055] The overall downgrade magnitude D is calculated based on a single event or a combination of events: For example, for a single event: B=2 levels, S=0.6 → D=ceil(1.2)=2 levels.

[0056] Composite events: Multiple events occur within a preset time window (e.g., 24 hours). First, calculate the independent degradation magnitude D_i for each event, and then calculate D_total. The α in the attenuation factor _k can be configured to 0.3. α is used to avoid excessive degradation caused by the superposition of multiple minor events.

[0057] The maximum downgrade limit is: no more than the total number of tiers from the current cycle to the lowest cycle (real-time deduction).

[0058] Example of downgrade level in the embodiment (starting from the end of the month): Tier 0: Monthly settlement; Tier 1: Semi-monthly settlement; Tier 2: Weekly settlement; Tier 3: Settlement by day; Tier 4: Real-time deduction.

[0059] In step 104, the settlement cycle is downgraded: the settlement cycle of the enterprise is adjusted by reducing the duration of the settlement cycle.

[0060] Update the airline's settlement cycle configuration immediately. The new cycle is equal to the original cycle downgraded to D level.

[0061] Figure 3This is another specific example diagram of the event-driven automatic settlement cycle adjustment method in the embodiments of the present invention, as shown in the figure. Figure 3 As shown, after adjusting the enterprise's settlement cycle in step 104 by utilizing the reduced settlement cycle duration, the method further includes: Step 301: Update the minimum balance warning threshold for prepayments and use the minimum balance warning threshold for warnings; or, send the settlement cycle adjustment to the enterprise, record the reason for the settlement cycle adjustment, and store it in the audit log.

[0062] If the settlement cycle after downgrading is 3-tier daily settlement or 4-tier real-time deduction, the minimum prepayment balance threshold will be increased simultaneously (e.g., by 50% to 100%).

[0063] Record the reasons for the downgrade (list of associated event IDs and timestamps) and save them in the audit log.

[0064] Airlines will be notified of the downgrade information and reasons through pre-set channels (App, SMS, email).

[0065] Step 105: Retrospective observation window and tiered recovery: Monitor the impact of risk events on the enterprise and the enterprise's credit rating within the retrospective observation window, and issue early warnings or tiered recovery of the settlement cycle based on the monitoring situation.

[0066] In one embodiment, the duration of the backtracking observation window is positively correlated with the number of gear reductions. That is, the deeper the gear reduction, the longer the backtracking observation window.

[0067] In one embodiment, the duration of the retrospective observation window is determined according to the number of reduction levels using the following formula: The observation window length W = base window × (D + 1) × event number correction factor; the base window is preset, for example, 7 days, D is D_i or D_total, and the event number correction factor is a multiplier factor positively correlated with the number of risk events involved in triggering this downgrade. It can be seen that the deeper the downgrade and the more events, the longer the observation period.

[0068] In this embodiment of the invention, the event quantity correction coefficient is a multiplier factor positively correlated with the number of risk events involved in triggering the downgrade (i.e., the total number of events when a compound event occurs). Its function is to extend the observation window. When multiple risk events (such as overdue payments, sudden increases in arrears, or rating downgrades) occur consecutively or simultaneously within a short period, even if the calculated composite downgrade level D is the same, the larger number of events indicates a broader and more complex risk exposure for the airline. Therefore, a longer observation period than for a single event is needed to verify whether the airline has truly recovered. This correction coefficient serves this purpose.

[0069] An example illustrating the function of the event quantity correction factor: Basic settings: Basic window = 7 days; Event quantity correction factor = 1 + (event quantity - 1) × 0.2 (increases linearly, increasing by 0.2 for each additional event).

[0070] Scenario 1: Single event degradation: An airline triggered a risk event, and the calculated downgrade level D = 2.

[0071] Event quantity correction factor = 1 + (1-1)×0.2 = 1.0; The observation window W = 7 days × (2+1) × 1.0 = 21 days.

[0072] Scenario 2: Two events occur simultaneously (or within a short period of time): If the same airline triggers two risk events within 24 hours (e.g., 3 days overdue + a sudden increase of 50% in outstanding payments), the overall downgrade level will still be D = 2 (calculated by the attenuation factor and rounded down).

[0073] Event quantity correction factor = 1 + (2-1)×0.2 = 1.2; The observation window W = 7 days × (2+1) × 1.2 = 25.2 days (rounded to 25 days).

[0074] Scenario 3: Three events occur: Even if three events are triggered (such as overdue payments, a sudden increase in debt, or an IATA rating downgrade), D will still be classified as level 2.

[0075] Event quantity correction factor = 1 + (3-1)×0.2 = 1.4; The observation window W = 7 days × 3 × 1.4 = 29.4 days (approximately 29 days).

[0076] For comparison details, please refer to Table 3 below: Table 3

[0077] The core logic is that even if the calculated downgrade level D is the same (both are level 2), the more events leading to the downgrade, the more the airline is exposed to multiple risk sources simultaneously, indicating lower stability. Therefore, a longer observation window is needed to confirm that the risks have truly subsided and to prevent a premature recovery. By using an event quantity correction factor, the length of the recovery period is determined not only by the depth of the downgrade but also by the breadth of the risks.

[0078] In one embodiment, monitoring the impact of risk events on an enterprise and its credit rating within a retrospective observation window, and issuing early warnings or tiered recovery of the settlement cycle based on the monitoring results, may include: Within the backtracking observation window, a tiered recovery settlement cycle will be performed when all of the following conditions are met: Condition 1: No new risk events have occurred within the retrospective observation window; Condition 2: The company's credit rating has not declined; Condition 3: The impact of the risk event that caused the settlement cycle to decrease returns to normal.

[0079] Specifically, the recovery conditions (which must be met simultaneously) are as follows: 1. No new triggering events occurred within the observation window; 2. The airline's regular risk score has remained stable in the core range of the pre-downgrade bracket (or has improved). 3. If the downgrade is triggered by a specific quantitative event (such as a sudden increase in debt), the indicator has recovered to below the normal threshold.

[0080] Step-by-step recovery: When all the above conditions are met, the window length is restored by one level. Each time, the window length is restored by one level. After each restoration, the observation window is recalculated and the window length is shortened step by step.

[0081] Let the initial overall degradation magnitude be D (calculated in step 103), and the number of recoveries be r (initially r=0). Then the window length after the r-th recovery (i.e., the next observation period) is: W = Base window × (D - r + 1) × Event count correction factor; The formula (D - r + 1) decreases with the number of recovery cycles, thus gradually shortening the window. When r = D, the airline has fully recovered to the original cycle and no longer needs to observe the window.

[0082] Quick rollback: If the triggering event occurs again during the observation period, the downgrade calculation will be re-executed (may be cumulative), that is, the reduction duration of the settlement cycle will be recalculated in step 103, and the observation period will be reset.

[0083] Furthermore, in one embodiment, the method further includes the fusion of event impact decay and a long-term credit model. Event-driven downgraded records are input as features into a periodic credit scoring model, but with time decay weights assigned. The periodic credit scoring model outputs the credit rating of each enterprise, and the basic settlement cycle of the enterprise is determined based on the credit rating, so that the basic settlement cycle of the enterprise can be adjusted in step 104 using the reduced duration of the settlement cycle.

[0084] The periodic credit scoring model is the foundational credit assessment module upon which this invention's embodiments rely. This model operates on a fixed schedule (typically monthly), taking into account structured data such as the airline's historical payment on-time rate, external credit ratings (e.g., IATA rating), current outstanding balance, aging distribution, and dispute rate. Using methods such as weighted scoring cards or logistic regression, it outputs a comprehensive credit score (e.g., 0-100 points) and a corresponding risk level (Excellent, Normal, Watchlist, High Risk, Severe Risk). This score is used to determine the airline's basic settlement cycle under normal circumstances (e.g., monthly, semi-monthly settlement) and serves as the basis for periodic adjustments to the cycle.

[0085] However, this model has an inherent lag. It cannot respond in real time to sudden risk events that occur between two assessment days, which is precisely the shortcoming that the event-driven mechanism of this invention aims to overcome. In this invention, event-driven downgrade records are input as short-term features into the periodic credit scoring model and given time decay weights, thereby achieving an organic integration of long-term credit and short-term events. This allows the model to "remember" recent risk events during the next periodic assessment without permanently damaging the airline's credit.

[0086] 0-30 days after the event: weight 1.0; 30–60 days: weight 0.5; 60–90 days: weight 0.2; More than 90 days: Weight 0; This ensures that short-term events do not permanently damage an airline's credit score.

[0087] Example 1: Degradation triggered by a single event.

[0088] 1. Initial conditions: Airline A's current settlement cycle is monthly (excellent risk rating). System configuration event E1: "IATA credit rating downgrade", B=2 levels, S=1.0.

[0089] 2. Event Occurrence: On a certain day, IATA downgraded an airline's rating from "A" to "BBB". The system detected this event in real time and triggered a reassessment. D = ceil(2×1.0) = 2 levels, downgrade from monthly settlement to weekly settlement by 2 levels.

[0090] 3. The system update cycle is weekly. A notification will be sent: "Due to the downgrade of your IATA rating, your settlement cycle has been temporarily adjusted to weekly, with a 21-day observation period."

[0091] Set the observation window W = 7 days × (2+1) × 1 = 21 days (using the formula W = base window × (D+1) × correction coefficient).

[0092] 4. Recovery process: After 21 days, the following conditions are met: no new events, stable periodic scores, and no further decline in rating. If these conditions are met, the rating will automatically be upgraded: weekly settlement → bi-weekly settlement (reverting to level 1).

[0093] New observation window: W = 7 days × (1+1) = 14 days.

[0094] If no abnormalities are found after 14 days, the process will be upgraded: from semi-monthly settlement to monthly settlement, resulting in full recovery.

[0095] Example 2: Compound event superposition and degradation.

[0096] 1. Initial conditions: Airline B's current cycle is semi-monthly. The following events occurred consecutively within 24 hours: Event E2 (overdue for 3 days): B=1 level, S=0.8 → D1=1 level; Event E3 (Sudden increase of 50% in debt): B=Level 1, S=0.9 → D2=Level 1.

[0097] 2. Downgrade calculation: Using a decay factor α=0.3, the event sequence is: E2(k=1, decay=1), E3(k=2, decay=e^{-0.3}=0.74); D_total = min(maximum downgrade limit 4, 1×1 + 1×0.74)=1.74 → rounded down to 2 levels; From semi-monthly settlement down 2 levels → daily settlement (semi-monthly settlement → weekly settlement down 1 level, weekly settlement → daily settlement down 1 level); The prepayment threshold has been increased by 50%.

[0098] 3. Recovery: Observation window W = 7 × (2+1) × 1.2 (event count adjustment) = 25 days. After the period expires, the ladder will be restored.

[0099] Example 3: Event debouncing and fast rollback.

[0100] 1. Airline C triggered the "overdue for more than 3 days" event, and the system downgraded it by one level (semi-monthly settlement → weekly settlement). Two hours later, the same airline triggered the same type of event again (still overdue). The system detected the repetition within the de-shaking window (1 hour) and ignored the second instance.

[0101] 2. On the 5th day of the observation period, Airline C again defaulted on its new bill, triggering a new event. The system recalculated: The current cycle is weekly (down 1 level), new event B=1 level, S=0.8 → D=1 level; From weekly settlement downgraded by one level → daily settlement; The observation period is reset to the new window length corresponding to daily settlement.

[0102] This invention also provides an event-driven automatic settlement cycle adjustment device, as described in the following embodiments. Since the principle by which this device solves the problem is similar to that of the event-driven automatic settlement cycle adjustment method, the implementation of this device can refer to the implementation of the event-driven automatic settlement cycle adjustment method; repeated details will not be elaborated further.

[0103] Figure 4 This is a schematic diagram of an event-driven automatic settlement cycle adjustment device according to an embodiment of the present invention, such as... Figure 4 As shown, the device 400 includes: The risk event subscription configuration module 401 is used to receive multiple risk events configured by the user through a visual interface; the attributes configured for each event include data source, severity, and base degradation level; the base degradation level reflects the reduction range of the settlement cycle when the event is triggered individually; The monitoring module 402 is used to detect risk events in real time through the data source, and generate a risk event record when a risk event is detected; the risk event record includes attribute information of a single event or a compound event; The instant reassessment module 403 is used to determine the reduction duration of the settlement cycle based on a preset formula, using the severity of each event in the risk event record and the base downgrade level; the preset formula is based on the base downgrade level, integrates the severity of the event, and uses a mathematical function to quantify the reduction duration of the settlement cycle. The cycle degradation execution module 404 is used to adjust the enterprise's settlement cycle by utilizing the reduced duration of the settlement cycle; The retrospective observation module 405 is used to monitor the impact of risk events on enterprises and their credit rating within the retrospective observation window, and to issue early warnings or gradually restore the settlement cycle based on the monitoring results.

[0104] In one embodiment, the risk event subscription configuration module 401 is specifically used for: The risk event category, event subtype, and event description can be configured by selecting from drop-down menus, checking boxes, or using custom input boxes. The severity of the event is determined by automatically recommending an initial severity value based on historical data statistics, rule template settings, or machine learning. The severity represents the weight of the event's contribution to the risk.

[0105] In one embodiment, the monitoring module 402 is specifically used for: Multi-source heterogeneous data is obtained in real time by polling various data sources and / or by subscribing to push data. Entity recognition, semantic feature extraction, and risk rule matching are performed on multi-source heterogeneous data through stream processing; the risk rules are obtained based on the configuration information provided by the user through a visual interface. When a match is successful, a risk event record is generated; the risk event record includes a timestamp, enterprise identifier, risk event type, severity, and base downgrade level.

[0106] In one embodiment, the instant reassessment module 403 is specifically used for: For a single event, Di is calculated using the following preset formula: D_i = ceil(B_i × S_i); where D_i represents the number of downgrade levels triggered independently by the i-th event, ceil() represents rounding up the result, B_i represents the severity coefficient, B_i∈[0,1], S_i represents the base downgrade level, S_i is any integer level between level 0 and level 4, where different levels correspond to different settlement periods of different time lengths; For composite events, D_total is calculated using the following preset formula: D_total = min(maximum downgrade limit, Σ(D_i × attenuation factor_k)); where D_total represents the number of downgrades, and the maximum downgrade limit represents the total number of downgrades from the current settlement period to the lowest settlement period. The attenuation factor _k = e^{-(k-1)×α}, where k is the sequence number of the event sorted by time (attenuation factor = 1 when k=1), and α is the configurable attenuation coefficient.

[0107] In one embodiment, the device 400 further includes: The auxiliary measures module is used to update the minimum balance warning threshold of prepayments after the periodic degradation execution module 404 adjusts the settlement cycle of the enterprise by reducing the settlement cycle duration, and to issue a warning using the minimum balance warning threshold; or, to send the settlement cycle adjustment information to the enterprise, record the reason for the settlement cycle adjustment, and store it in the audit log.

[0108] In one embodiment, the duration of the retrospective observation window is positively correlated with the number of reduction levels.

[0109] In one embodiment, the duration of the retrospective observation window is determined according to the number of reduction levels using the following formula: The observation window length W = base window × (D + 1) × event number correction factor; the base window is preset, D is D_i or D_total, and the event number correction factor is a multiplier factor that is positively correlated with the number of risk events involved in triggering this downgrade.

[0110] In one embodiment, the backtracking observation module 405 is specifically used for: Within the backtracking observation window, a tiered recovery settlement cycle will be performed when all of the following conditions are met: Condition 1: No new risk events have occurred within the retrospective observation window; Condition 2: The company's credit rating has not declined; Condition 3: The impact of the risk event that caused the settlement cycle to decrease returns to normal.

[0111] This invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the above-described event-driven automatic adjustment method for settlement cycles.

[0112] Figure 5 This is a schematic diagram of a computer device in an embodiment of the present invention, such as... Figure 5 As shown, this embodiment of the invention also provides a computer device 500, including a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program 503, it implements the above-mentioned event-driven automatic adjustment method for settlement cycle.

[0113] This invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned event-driven automatic settlement cycle adjustment method.

[0114] This invention also provides a computer program product, which includes a computer program that, when executed by a processor, implements the above-mentioned event-driven automatic adjustment method for settlement cycles.

[0115] The embodiments of the present invention have the following beneficial effects: Real-time performance: Response within seconds after a serious risk event occurs, immediate adjustment of the settlement cycle, shortening the risk exposure window from weeks to seconds, and significantly reducing the risk of bad debts.

[0116] Refined downgrade: The downgrade magnitude is adaptively calculated based on the severity coefficient of the event and the basic downgrade level, avoiding a "one-size-fits-all" approach and balancing risk control and business continuity.

[0117] Flexible and configurable: Risk control personnel can dynamically add or delete event types, adjust thresholds and degradation parameters without modifying the code, adapting to the risk preferences of different airports.

[0118] Complex event handling: By using attenuation factors to reasonably superimpose the effects of multiple events, risks can be avoided from being over-degraded or overlooked.

[0119] Automated recovery: Automatic step-by-step recovery after the observation period reduces manual maintenance costs. The recovery process is linked to the severity of the event to prevent premature rebound.

[0120] Events are integrated with the credit model: the impact of events decays over time, and short-term events will not permanently drag down credit scores, reflecting fairness.

[0121] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0123] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0125] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An event-driven method for automatically adjusting the settlement cycle, characterized in that, include: Receive multiple risk events configured by the user through a visual interface; Each event is configured with properties including data source, severity, and base degradation level. The number of base downgrade levels reflects the range of reduction in the settlement cycle when an event is triggered individually; Risk events are detected in real time through the data source, and a risk event record is generated when a risk event is detected. The risk event record includes attribute information for single events or compound events; Based on a preset formula, the reduction duration of the settlement cycle is determined by utilizing the severity of each event in the risk event record and the base downgrade level. The preset formula is based on the number of basic downgrade levels, takes into account the severity of the event, and uses mathematical functions to quantify the reduction in the settlement cycle. The reduced settlement cycle duration is used to adjust the company's settlement cycle. Within the retrospective observation window, monitor the impact of risk events on enterprises and their credit ratings, and issue early warnings or gradually restore settlement cycles based on the monitoring results.

2. The method as described in claim 1, characterized in that, Receives multiple risk events configured by the user through a visual interface, including: The risk event category, event subtype, and event description can be configured by selecting from drop-down menus, checking boxes, or using custom input boxes. The severity of the event is determined by automatically recommending an initial severity value based on historical data statistics, rule template settings, or machine learning. The severity represents the weight of the event's contribution to the risk.

3. The method as described in claim 1, characterized in that, Risk events are detected in real time through the data source. When a risk event is detected, a risk event record is generated, including: Multi-source heterogeneous data is obtained in real time by polling various data sources and / or by subscribing to push data. Entity recognition, semantic feature extraction, and risk rule matching are performed on multi-source heterogeneous data through stream processing; the risk rules are obtained based on the configuration information provided by the user through a visual interface. When a match is successful, a risk event record is generated; the risk event record includes a timestamp, enterprise identifier, risk event type, severity, and base downgrade level.

4. The method as described in claim 1, characterized in that, Based on a preset formula, the reduction in settlement cycle duration is determined using the severity of each event in the risk event record and the base downgrade level, including: For a single event, Di is calculated using the following preset formula: D_i = ceil(B_i × S_i); where D_i represents the number of downgrade levels triggered independently by the i-th event, ceil() represents rounding up the result, B_i represents the severity coefficient, B_i∈[0,1], S_i represents the base downgrade level, S_i is any integer level between level 0 and level 4, where different levels correspond to different settlement periods of different time lengths; For composite events, D_total is calculated using the following preset formula: D_total = min(maximum downgrade limit, Σ(D_i × attenuation factor_k)); where D_total represents the number of downgrades, and the maximum downgrade limit represents the total number of downgrades from the current settlement period to the lowest settlement period. The attenuation factor _k = e^{-(k-1)×α}, where k is the sequence number of the event sorted by time (attenuation factor = 1 when k=1), and α is the configurable attenuation coefficient.

5. The method as described in claim 1, characterized in that, After adjusting the company's settlement cycle by utilizing the reduced settlement cycle duration, the process also includes: Update the minimum balance warning threshold for prepayments and use the minimum balance warning threshold for warning purposes; Alternatively, the settlement cycle adjustment information can be sent to the enterprise, the reasons for the settlement cycle adjustment can be recorded, and the information can be stored in the audit log.

6. The method as described in claim 4, characterized in that, The duration of the retrospective observation window is positively correlated with the number of gear reductions.

7. The method as described in claim 6, characterized in that, The duration of the retrospective observation window is determined according to the number of reductions in the gear ratio using the following formula: The observation window length W = base window × (D + 1) × event number correction factor; the base window is preset, and D is D_i or D_total; the event number correction factor is a multiplier factor that is positively correlated with the number of risk events involved in triggering this downgrade.

8. The method as described in claim 1, characterized in that, Within the retrospective observation window, monitor the impact of risk events on enterprises and their credit ratings, and issue early warnings or implement tiered recovery of settlement cycles based on the monitoring results, including: Within the backtracking observation window, a tiered recovery settlement cycle will be performed when all of the following conditions are met: Condition 1: No new risk events have occurred within the retrospective observation window; Condition 2: The company's credit rating has not declined; Condition 3: The impact of the risk event that caused the settlement cycle to decrease returns to normal.

9. An event-driven automatic settlement cycle adjustment device, characterized in that, include: The risk event subscription configuration module is used to receive multiple risk events configured by the user through a visual interface; Each event is configured with properties including data source, severity, and base degradation level. The number of base downgrade levels reflects the range of reduction in the settlement cycle when an event is triggered individually; The monitoring module is used to detect risk events in real time through the data source, and generate risk event records when a risk event is detected. The risk event record includes attribute information for single events or compound events; The instant reassessment module is used to determine the reduction duration of the settlement cycle based on a preset formula, using the severity of each event in the risk event record and the base downgrade level. The preset formula is based on the number of basic downgrade levels, takes into account the severity of the event, and uses mathematical functions to quantify the reduction in the settlement cycle. The periodic degradation execution module is used to adjust the enterprise's settlement cycle by utilizing the reduced duration of the settlement cycle; The retrospective observation module is used to monitor the impact of risk events on enterprises and their credit rating within the retrospective observation window, and to issue early warnings or gradually restore the settlement cycle based on the monitoring results.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method of any one of claims 1 to 8.

12. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the method of any one of claims 1 to 8.

Images