Data processing method and device, electronic equipment and storage medium

Abstract

The present disclosure relates to the technical field of computers, and relates to a data processing method and device, electronic equipment and storage medium. The data processing comprises: obtaining a service data set of a first subject, the service data set comprising first service data of an uncompleted first service between the first subject and a second subject, a service quantity of at least one historical service between the first subject and the second subject, and at least one set of historical service data of the at least one historical service; predicting a service duration of the first service based on at least one of the service quantity and the at least one set of historical service data, and a service start time in the first service data; predicting an end time of the first service based on the service start time and the service duration; and predicting a service risk of the first subject based on end times of all uncompleted services between the first subject and all subjects in a service relationship with the first subject.

Description

Technical Field

[0001] This disclosure relates to the field of computer technology, and more specifically, to a data processing method, apparatus, electronic device, and storage medium. Background Technology

[0002] Currently, business management tools are widely used in various economic organizations to improve the efficiency of business decision-making, optimize resource allocation, and reduce business risks.

[0003] However, existing business management tools have a wide variety of data sources, which may differ in format, storage method, update frequency, or collection standards. Therefore, existing business management tools need to use complex data architectures to process a large amount of business data in order to effectively manage the business data, which consumes a lot of computing resources.

[0004] It should be noted that the information in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this disclosure is to provide a data processing method, apparatus, electronic device, and storage medium, thereby overcoming, to at least a certain extent, the deficiencies of the related technologies and effectively saving computing resources.

[0006] Other features and advantages of this disclosure will become apparent from the following detailed description, or may be learned in part from practice of this disclosure.

[0007] According to a first aspect of this disclosure, a data processing method is provided, the method comprising: Obtain the business dataset of the first entity, the business dataset including the first business data of the first business between the first entity and the second entity that has not been completed, the business quantity of at least one historical business between the first entity and the second entity, and at least one set of historical business data of the at least one historical business, the second entity being any entity that has a business relationship with the first entity, and the first business being any business between the first entity and the second entity that has not been completed; Based on the number of services and at least one item from the at least one set of historical service data, as well as the service start time in the first service data, the duration of the first service is predicted. Based on the start time and duration of the service, predict the end time of the first service; Based on the end time of all unfinished business between the first entity and all entities with which it has business relationships, business risk is predicted for the first entity.

[0008] In some embodiments of this disclosure, predicting the duration of the first service based on the number of services and at least one of the at least one set of historical service data, and the service start time in the first service data, includes: If the number of services is less than or equal to a first preset number, then based on the industry characteristic coefficient of the industry in which the first entity operates and the industry benchmark service duration of the industry in which the first entity operates, the duration of the first service is determined, and the duration of the first service is determined as the service duration of the first service; or, If the number of services is greater than or equal to the second preset number, the duration of the second service is determined based on the at least one set of historical service data. Based on the duration of the second service and the attribute information of the second entity, the duration of the first service is determined; or, If the number of services is greater than the first preset number and less than the second preset number, then the duration of the first service is determined based on the duration of the first service, the duration of the second service, and the attribute information. The second preset quantity is greater than the first preset quantity.

[0009] In some embodiments of this disclosure, each set of historical business data includes the start time, end time, and amount of the corresponding historical business; determining the duration of the second business based on the at least one set of historical business data includes: The actual duration of each historical service is determined based on its start and end times. Calculate the first ratio between the transaction amount of each historical transaction and the actual duration of that historical transaction; Calculate the difference between the actual duration of each historical service and the predicted duration of that historical service, and the second ratio between the difference in duration of that historical service and the predicted duration of that historical service. The duration of the second service is determined based on the sum of at least one first ratio corresponding to the at least one historical service, the service amount of the first service, and at least one second ratio corresponding to the at least one historical service.

[0010] In some embodiments of this disclosure, determining the duration of the first service based on the duration of the second service and the attribute information of the second subject includes: Based on the attribute information of the second subject, the service type of the first service, the at least one set of historical service data, and the duration of the second service, a first anomaly correction score is determined; The duration of the second service is corrected using the first anomaly correction score to obtain the duration of the first service.

[0011] In some embodiments of this disclosure, the attribute information includes the subject type of the second subject and the comment information of the second subject. Determining the first anomaly correction score based on the attribute information of the second subject, the service type of the first service, the at least one set of historical service data, and the duration of the second service includes: Based on the subject type and the business type, determine the business anomaly score of the first business; Based on the statistical value of at least one actual duration of the at least one historical service and the duration of the second service, determine the duration anomaly score of the first service. Based on at least one transaction amount of the at least one historical transaction and the transaction amount of the first transaction, determine the abnormal transaction amount score of the first transaction; Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. The abnormal business score, the abnormal industry score, the abnormal duration score, and the abnormal business amount score are weighted and calculated to obtain a comprehensive abnormal score. The first abnormality correction score is obtained based on the comprehensive abnormality score, the benchmark abnormality score, and the preset abnormality correction coefficient.

[0012] In some embodiments of this disclosure, the method further includes: If the overall anomaly score is greater than the first preset score and less than or equal to the second preset score, the duration of the first service is adjusted based on the service amount of the first service, the average service amount of the at least one historical service, and a preset amount correction coefficient. The second preset score is greater than the first preset score.

[0013] In some embodiments of this disclosure, the method further includes: If the first condition is met, a prompt message is displayed, which is used to indicate that the first service is abnormal; The first condition includes at least one of the following: The overall anomaly score is greater than or equal to the aforementioned second preset score; The average of N duration deviations is greater than or equal to a first preset threshold. The N duration deviations include the deviation between the predicted duration of each historical service in N consecutive historical services and the actual duration of that historical service. The attribute information of the second subject indicates that the second subject is at risk; The first business data indicates that the first business is at risk.

[0014] In some embodiments of this disclosure, before predicting the business risk of the first entity based on the end time of all outstanding business transactions between the first entity and all entities with which it has a business relationship, the method further includes: Based on the service duration of the first service, the comprehensive anomaly score, at least one actual duration of the at least one historical service, and at least one predicted duration of the at least one historical service, the service duration interval in which the service duration of the first service falls is determined. Based on the duration range of the service and the start time of the service, the time range in which the end time falls is determined; The business risk prediction for the first entity based on the end times of all outstanding business transactions between the first entity and all entities with which it has business relationships includes: Based on the end time of all unfinished business between the first entity and all entities with which it has a business relationship, the time interval in which the end time of each unfinished business falls, the type of each unfinished business, and the entity type of the entity corresponding to each unfinished business, business risk prediction is performed on the first entity.

[0015] In some embodiments of this disclosure, the attribute information includes the comment information of the second subject and the subject type of the second subject. Determining the duration of the first service based on the duration of the first service, the duration of the second service, and the attribute information of the second subject includes: Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. Based on the industry anomaly score and the second anomaly correction score, a first weight value for the duration of the first business and a second weight value for the duration of the second business are determined. The first weight value and the second weight value are used to perform a weighted calculation on the duration of the first service and the duration of the second service to obtain a weighted duration. A second anomaly correction score is determined based on the industry anomaly score, the subject type, the business type, the at least one set of historical business data, and the weighted duration. The weighted duration is corrected using the second anomaly correction score to obtain the service duration of the first service.

[0016] In some embodiments of this disclosure, determining the first weight value of the first service duration and the second weight value of the second service duration based on the industry anomaly score and the second anomaly correction score includes: The second weight value is determined based on the industry anomaly score; The first weight value is determined based on the second weight value and the second anomaly correction score.

[0017] According to a second aspect of this disclosure, a data processing apparatus is provided, the apparatus comprising: The acquisition module is used to acquire the business dataset of the first subject. The business dataset includes the first business data of the first business that has not been completed between the first subject and the second subject, the business quantity of at least one historical business between the first subject and the second subject, and at least one set of historical business data of the at least one historical business. The second subject is any subject that has a business relationship with the first subject, and the first business is any business that has not been completed between the first subject and the second subject. The prediction module is used to predict the duration of the first service based on the number of services and at least one item from the at least one set of historical service data, as well as the service start time in the first service data. The prediction module is also used to predict the end time of the first service based on the service start time and the service duration. The prediction module is also used to predict business risks for the first entity based on the end time of all unfinished business between the first entity and all entities with which it has business relationships.

[0018] According to a third aspect of this disclosure, an electronic device is provided, including a processor; Memory used to store processor-executable instructions; The processor is configured as follows: The steps to implement the method described in the first aspect.

[0019] According to a fourth aspect of the present disclosure, a non-transitory computer-readable storage medium is provided, which, when instructions in the storage medium are executed by a processor of a mobile terminal, enables the mobile terminal to perform the data processing method described in the first aspect.

[0020] According to a fifth aspect of the present disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the data processing method as described in the first aspect.

[0021] The data processing method in the exemplary embodiments of this disclosure involves acquiring a business dataset of a first entity. The business dataset includes first business data of unfinished first business between the first entity and a second entity, the number of business transactions of at least one historical business between the first entity and the second entity, and at least one set of historical business data for the at least one historical business. The second entity is any entity with a business relationship with the first entity, and the first business is any unfinished business between the first entity and the second entity. Based on the number of business transactions, at least one item from the at least one set of historical business data, and the business start time in the first business data, the method predicts the duration of the first business transaction. Based on the business start time and the duration of the business transaction, the method predicts the end time of the first business transaction. Based on the end times of all unfinished business transactions between the first entity and all entities with which it has a business relationship, the method performs business risk prediction for the first entity. Thus, the data processing method provided in this disclosure can predict the end time of an incomplete business by using any business data of the first entity's unfinished business in the first entity's business dataset, as well as historical business data between the first entity and the second entity corresponding to the unfinished business. After the electronic device predicts the end time of all unfinished businesses between the first entity and all entities with which it has business relationships, it can perform business risk prediction for the first entity based on the end time of all unfinished businesses between the first entity and all entities with which it has business relationships. As a result, the electronic device only needs to perform calculations on a small amount of data in the first entity's business dataset to perform business risk prediction for the first entity, thereby effectively saving computing resources and improving the efficiency of risk prediction.

[0022] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0023] The above and other objects, features, and advantages of this disclosure will become readily apparent from the following detailed description of exemplary embodiments, taken in conjunction with the accompanying drawings. Several embodiments of this disclosure are illustrated in the drawings by way of example and not limitation.

[0024] Figure 1 A flowchart illustrating a data processing method according to an exemplary embodiment of the present disclosure is shown.

[0025] Figure 2 A schematic diagram of the overall architecture of a data processing system according to an exemplary embodiment of the present disclosure is shown.

[0026] Figure 3 A schematic diagram illustrating the workflow of a dual-mode prediction module according to an exemplary embodiment of the present disclosure is shown.

[0027] Figure 4 A schematic diagram of the structure of an exception handling module according to an exemplary embodiment of the present disclosure is shown.

[0028] Figure 5 A funds level chart in an application-layer "management dashboard" of an exemplary embodiment of this disclosure is shown.

[0029] Figure 6 A schematic diagram of the parameter optimization process of a feedback mechanism according to an exemplary embodiment of the present disclosure is shown.

[0030] Figure 7 An overall timing diagram of the processing flow of a data processing system according to an exemplary embodiment of the present disclosure is shown.

[0031] Figure 8 A block diagram of a data processing apparatus according to an exemplary embodiment of the present disclosure is shown.

[0032] Figure 9 A schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present disclosure is shown.

[0033] In the accompanying drawings, the same or corresponding reference numerals indicate the same or corresponding parts. Detailed Implementation

[0034] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar structures, and therefore their detailed description will be omitted.

[0035] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced without one or more of the specific details described, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known structures, methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0036] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, or in one or more software-hardened modules, or in different network and / or processor devices and / or microcontroller devices.

[0037] refer to Figure 1 The diagram shown is a flowchart of a data processing method according to an exemplary embodiment of this disclosure, which can be applied to an electronic device. Figure 1 As shown, the data processing method may include the following steps 101 to 104: Step 101: The electronic device acquires the business dataset of the first entity.

[0038] In some embodiments of this disclosure, the aforementioned business dataset may include first business data of an unfinished first business between the first entity and the second entity, the business quantity of at least one historical business between the first entity and the second entity, and at least one set of historical business data of the at least one historical business. The second entity may be any entity that has a business relationship with the first entity, and the first business may be any unfinished business between the first entity and the second entity.

[0039] In some embodiments of this disclosure, the electronic device can acquire the business data of the first subject if the conditions for acquiring the business dataset are met.

[0040] In some embodiments of this disclosure, the above-mentioned business data acquisition conditions include any one of the following: The preset time interval has elapsed since the last acquisition of the business dataset. The system receives a user's instruction to retrieve business data, which includes the aforementioned first business data.

[0041] In some embodiments of this disclosure, a preset time interval is reached after the last acquisition of business data, indicating that the electronic device can periodically collect the business data set of the first subject according to the preset time interval.

[0042] In some embodiments of this disclosure, receiving a user's business data acquisition instruction can be understood as the electronic device acquiring the business dataset of the first entity whenever the user uploads business data of a new business of the first entity and the second entity to the electronic device.

[0043] Step 102: The electronic device predicts the duration of the first service based on the number of services and at least one item from at least one set of historical service data, as well as the service start time in the first service data.

[0044] In some embodiments of this disclosure, step 102 can be implemented by step 102a, or by steps 102b and 102c, or by step 102d: Step 102a: If the number of services is less than or equal to the first preset number, the electronic device determines the duration of the first service based on the industry characteristic coefficient of the industry in which the first entity is located and the industry benchmark service duration of the industry in which the first entity is located, and determines the duration of the first service as the duration of the first service.

[0045] In some embodiments of this disclosure, the first preset quantity can be understood as a preset minimum number of services.

[0046] In some embodiments of this disclosure, the number of services is less than or equal to the first preset number, indicating that the number of historical services between the first entity and the second entity is very small. Therefore, at least one set of historical service data for at least one historical service between the first entity and the second entity is insufficient as a basis for predicting the duration of the first service. Consequently, the electronic device can determine the duration of the first service based on the industry characteristic coefficient of the industry in which the first entity is located and the industry benchmark duration of the industry in which the first entity is located as the duration of the first service.

[0047] In some embodiments of this disclosure, the aforementioned first service data may further include the prepayment ratio of the service amount of the aforementioned first service. The electronic device may determine the duration of the aforementioned first service based on the aforementioned industry characteristic coefficient, the aforementioned industry benchmark service duration, and the aforementioned prepayment adjustment coefficient of the service amount of the aforementioned first service. In some embodiments of this disclosure, the electronic device may use the following calculation formula to determine the duration of the first service: (1) in, This indicates the duration of the first service mentioned above. This indicates the duration of the aforementioned industry benchmark business. This represents the adjustment coefficient for the new subject, which is dimensionless. It takes a value of 1.2 when the subject type of the first subject is "new subject", and otherwise takes a value of 1.0. The subject type of the first subject is "new subject". This indicates the aforementioned prepayment adjustment coefficient. =1-0.7r / 100, where r is the aforementioned prepayment ratio, with a value ranging from 0 to 100; This represents the acceleration factor for small amounts, which is dimensionless. When the amount of the first business transaction mentioned above is less than a preset threshold (such as 100,000 yuan), the value is 0.8; otherwise, the value is 1.0. This represents an industry characteristic coefficient, which is dimensionless and is adjusted according to the specific characteristics of different industries. The default value is 1.0.

[0048] It should be noted that the subject type of the first subject mentioned above is "new subject", which means that the second subject mentioned above is a new subject that has a business relationship with the first subject mentioned above. In other words, the number of historical business transactions between the first subject and the second subject mentioned above is 0, and the first business transaction mentioned above is the first business transaction between the first subject and the second subject mentioned above.

[0049] Step 102b: If the number of services is greater than or equal to the second preset number, the electronic device determines the duration of the second service based on at least one set of historical service data.

[0050] In some embodiments of this disclosure, the second preset quantity is greater than the first preset quantity.

[0051] In some embodiments of this disclosure, the aforementioned second preset quantity can be understood as a preset maximum number of services.

[0052] In some embodiments of this disclosure, the number of services is greater than or equal to the second preset number, indicating that there are many historical services between the first entity and the second entity. Therefore, at least one set of historical service data for at least one historical service between the first entity and the second entity is sufficient as a basis for predicting the duration of the first service. Thus, the electronic device can determine the duration of the first service by using the second service duration determined based on at least one set of historical service data and the attribute information of the second entity.

[0053] In some embodiments of this disclosure, each set of historical business data includes the start time, end time, and amount of the corresponding historical business; the step 102b above, "determining the duration of the second business based on at least one set of historical business data," can be implemented through the following steps 102b1 to 102b4: Step 102b1: The electronic device determines the actual duration of each historical service based on its start and end times.

[0054] In some embodiments of this disclosure, the electronic device can determine the actual duration of the historical service as the length of time between the service start time and the service start time of each historical service.

[0055] For example, if historical service A started on December 31, 2025, and ended on February 25, 2026, the time between December 31, 2025, and February 25, 2026 is 56 days. Therefore, the electronic device can determine that the actual duration of historical service A is 56 days.

[0056] Step 102b2: The electronic device calculates a first ratio between the transaction amount of each historical transaction and the actual duration of that historical transaction.

[0057] In some embodiments of this disclosure, the first ratio mentioned above can be understood as the transaction amount of the historical business within a unit of time.

[0058] Step 102b3: The electronic device calculates the difference between the actual duration of each historical service and the predicted duration of the historical service, as well as the second ratio between the difference in duration of the historical service and the predicted duration of the historical service.

[0059] In some embodiments of this disclosure, the electronic device can calculate the aforementioned duration difference using the following formula: (2) in, This indicates the time difference mentioned above. This represents the actual duration of the j-th historical transaction. This represents the predicted duration of the j-th historical business item mentioned above, where j is a positive integer.

[0060] In some embodiments of this disclosure, the electronic device may calculate the aforementioned second ratio using the following formula: (3) in, This represents the second ratio mentioned above.

[0061] Step 102b4: The electronic device determines the duration of the second service based on at least one first ratio corresponding to at least one historical service, the service amount of the first service, and the sum of at least one second ratio corresponding to at least one historical service.

[0062] In some embodiments of this disclosure, the electronic device may first perform a weighted operation on the at least one first ratio, then calculate the ratio of the weighted operation result of the at least one first ratio to the sum of at least one weight of the at least one first ratio, and then determine the duration based on the third ratio of the weighted operation result of the at least one first ratio to the sum of at least one weight of the at least one first ratio, the sum of the at least one second ratio, and the service amount of the first service.

[0063] In some embodiments of this disclosure, the weight of each of the at least one first ratio can be the time weight of the corresponding historical service.

[0064] In some embodiments of this disclosure, the electronic device may use the following formula to determine the weight of the first ratio: = (4) in, w j This represents the weight of the first ratio corresponding to the j-th historical business item mentioned above. This indicates the start time of the j-th historical service item mentioned above. This indicates the current time, which is the time when the aforementioned business dataset was obtained.

[0065] In some embodiments of this disclosure, the electronic device may calculate the aforementioned third ratio using the following formula: (5) in, The third ratio mentioned above can represent the weighted average amount / period ratio of at least one historical business, in yuan / day, where n represents the number of business transactions of at least one historical business.

[0066] In some embodiments of this disclosure, the electronic device may calculate the sum of at least one of the above-mentioned second ratios using the following formula: (6) in, The sum of the above-mentioned at least second ratios can represent the deviation of the prediction results of at least one of the above-mentioned historical business.

[0067] In some embodiments of this disclosure, the electronic device may use the following formula to calculate the duration of the second service: (7) in, This indicates the duration of the first service mentioned above. This indicates the business amount of the first business mentioned above.

[0068] Step 102c: The electronic device determines the duration of the first service based on the duration of the second service and the attribute information of the second subject.

[0069] In some embodiments of this disclosure, step 102c above can be implemented by the following steps 102c1 and 102c2: Step 102c1: The electronic device determines the first anomaly correction score based on the attribute information of the second subject, the service type of the first service, at least one set of historical service data, and the duration of the second service.

[0070] In some embodiments of this disclosure, the attribute information may include the subject type of the second subject and the comment information of the second subject, and step 102c1 may be implemented by the following steps A1 to A6: Step A1: Based on the subject type and service type, the electronic device determines the service anomaly score of the first service.

[0071] In some embodiments of this disclosure, when the subject type of the first entity is a new entity and the service type of the first service is a new service type, the electronic device can set the service anomaly score of the first service to a first preset score, such as 50. If there is a serious service anomaly marker in the first service data, such as the risk marker of the second entity, and the subject type of the first entity is a new entity and the service type of the first service is a new service type, the electronic device can set the service anomaly score of the first service to a second preset score, such as 100. Otherwise, the electronic device can set the service anomaly score to 0.

[0072] Step A2: The electronic device determines the duration anomaly score of the first service based on the statistical value of at least one actual duration of at least one historical service and the duration of the second service.

[0073] In some embodiments of this disclosure, the statistical value of the at least one actual duration may include the average of the at least one actual duration and the standard deviation of the at least one actual duration.

[0074] In some embodiments of this disclosure, the electronic device may first determine the first duration anomaly Z-score of the first service based on the average value of the at least one actual duration, the standard deviation of the at least one actual duration, and the duration of the second service, and then normalize the first duration anomaly Z-score to obtain the first duration anomaly score of the first service.

[0075] In some embodiments of this disclosure, the electronic device may calculate the aforementioned first duration anomaly Z-score using the following formula: (8) in, This indicates the Z-score representing the first duration anomaly mentioned above. This represents the average value mentioned above. This represents the standard deviation mentioned above.

[0076] In some embodiments of this disclosure, the electronic device may normalize the Z-score of the aforementioned first duration anomaly using the following formula: (9) in, This indicates the above-mentioned abnormal score for the first duration.

[0077] Step A3: The electronic device determines the abnormal score of the first service based on at least one service amount of at least one historical service and the service amount of the first service.

[0078] In some embodiments of this disclosure, the electronic device may first obtain the 25th percentile and the 75th percentile of the at least one business amount, then determine the abnormal multiple of the first business amount based on the 25th percentile, the 75th percentile and the business amount of the first business, and then normalize the abnormal multiple of the first business amount to obtain the abnormal score of the first business amount.

[0079] In some embodiments of this disclosure, the electronic device may use the following formula to calculate the aforementioned abnormal multiple of the amount: ； (10) in, This indicates an abnormal multiple of the aforementioned amount. This represents the 25th quantile mentioned above. This represents the 75th percentile mentioned above.

[0080] In some embodiments of this disclosure, the electronic device may normalize the aforementioned abnormal multiples of amount using the following formula: (11) in, This indicates an abnormal score for the above-mentioned business amount.

[0081] Step A4: Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the electronic device determines the industry anomaly score of the industry in which the second subject is located.

[0082] In some embodiments of this disclosure, the electronic device can determine the risk coefficient of the second subject based on the above-mentioned comment information, and obtain the industry anomaly score by weighting the risk coefficient of the second subject, the industry prosperity index, and the industry risk index. The value range of the industry anomaly score is 0 to 100.

[0083] Step A5: The electronic device performs a weighted calculation on the abnormal business score, abnormal industry score, abnormal duration score, and abnormal business amount score to obtain a comprehensive abnormal score.

[0084] In some embodiments of this disclosure, the electronic device may use the following formula to weight the above-mentioned business anomaly score, the above-mentioned industry anomaly score, the above-mentioned first duration anomaly score, and the above-mentioned business amount anomaly score to obtain a first comprehensive anomaly score: (12) in, This represents the first comprehensive anomaly score mentioned above, with a value range of 0 to 100; This indicates the abnormal score for the above-mentioned business amount; This indicates the abnormal scores for the aforementioned industries; , as well as These represent the weights of the above-mentioned first duration abnormal score, the above-mentioned business amount abnormal score, the above-mentioned business amount abnormal score, and the above-mentioned industry abnormal score, respectively, with default values ​​of 0.4, 0.3, 0.2, and 0.1.

[0085] Step A6: The electronic device obtains the first anomaly correction score based on the comprehensive anomaly score, the benchmark anomaly score, and the preset anomaly correction coefficient.

[0086] In some embodiments of this disclosure, the electronic device can calculate the difference between the first comprehensive anomaly score and the preset anomaly benchmark value, and then determine the first anomaly correction score based on the difference and the preset anomaly correction coefficient.

[0087] In some embodiments of this disclosure, the electronic device may use the following formula to calculate the aforementioned first anomaly correction score: (13) in, This indicates the correction score for the first anomaly mentioned above; This represents the preset anomaly correction coefficient, with a default value of 0.2. It is used to control the degree of anomaly correction on the duration of the second service mentioned above. This indicates the preset abnormal baseline value, which is 30 by default. When the overall abnormal score is lower than the preset abnormal baseline value, it means that the duration of the second service can be accurately used as the duration of the first service without correction.

[0088] Step 102c2: The electronic device uses the first anomaly correction score to correct the duration of the second service, thereby obtaining the duration of the first service.

[0089] In some embodiments of this disclosure, the electronic device may use the following formula to correct the duration of the second service described above: (14) in, This indicates the duration of the first service mentioned above. It can be seen from formula (13) that when... > hour, >1. By appropriately extending the duration of the second service mentioned above, the duration of the first service mentioned above can be predicted; when < hour, <1, by appropriately shortening the duration of the second service mentioned above, the duration of the first service mentioned above can be predicted; when = hour, =1, the duration of the second service mentioned above is the predicted duration of the first service.

[0090] Step 102d: If the number of services is greater than the first preset number and less than the second preset number, the electronic device determines the duration of the first service based on the duration of the first service, the duration of the second service, and attribute information.

[0091] In some embodiments of this disclosure, the attribute information includes the comment information of the second subject and the subject type of the second subject. Step 102d can be implemented by the following steps 102d1 to 102d5: Step 102d1: Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the electronic device determines the industry anomaly score of the industry in which the second subject is located.

[0092] It should be noted that the implementation process of step 102d1 above can refer to the relevant description in step A4 above. To avoid repetition, this embodiment will not repeat it here.

[0093] Step 102d2: Based on the industry anomaly score, the electronic device determines the first weight value of the duration of the first business and the second weight value of the duration of the second business.

[0094] In some embodiments of this disclosure, the electronic device determines the second weight value based on the aforementioned number of services, the aforementioned first preset number, the aforementioned second preset number, and the aforementioned industry anomaly score.

[0095] In some embodiments of this disclosure, the electronic device may use the following formula to determine the second weight mentioned above: (16) in This represents the second weight mentioned above, with a value ranging from 0 to 1. This indicates the first preset quantity mentioned above. This indicates the second preset quantity mentioned above. This represents the industry sensitivity coefficient, with a default value of 0.5.

[0096] In some embodiments of this disclosure, the electronic device may determine the difference between 1 and the second weight as the first weight.

[0097] Step 102d3: The electronic device uses the first weight value and the second weight value to perform a weighted calculation on the duration of the first service and the duration of the second service to obtain the weighted duration.

[0098] In some embodiments of this disclosure, the electronic device may use the following formula to weight the duration of the first service and the duration of the second service: (17) in, This indicates the weighted duration mentioned above.

[0099] Step 102d4: The electronic device determines the second anomaly correction score based on the industry anomaly score, subject type, business type, at least one set of historical business data, and weighted duration.

[0100] In some embodiments of this disclosure, the electronic device can determine a second duration anomaly score for the first service based on a statistical value of at least one actual duration of at least one historical service and the weighted duration mentioned above. Then, based on the industry anomaly score, the second duration anomaly score, the service anomaly score determined by the electronic device based on the subject type and the service type, and the service amount anomaly score determined by the electronic device based on at least one service amount of at least one historical service and the service amount of the first service, a second comprehensive anomaly score is determined. Finally, based on the second comprehensive anomaly score, the baseline anomaly score, and the preset anomaly correction coefficient, a second anomaly correction score is obtained.

[0101] It should be noted that the process by which the electronic device determines the abnormal score of the second duration based on the statistical value of at least one actual duration of at least one historical service and the weighted duration mentioned above can be referred to the relevant description in step A2 above. To avoid repetition, this embodiment will not repeat it here.

[0102] It should be noted that the process of determining the above-mentioned second comprehensive abnormal score based on the above-mentioned industry abnormal score, the above-mentioned second duration abnormal score, the above-mentioned business abnormal score determined by the electronic device based on the above-mentioned subject type and the above-mentioned business type, and the above-mentioned business amount abnormal score determined by the electronic device based on at least one business amount of at least one historical business and the business amount of the above-mentioned first business can refer to the relevant description in step A5 above. To avoid repetition, this embodiment will not repeat it here.

[0103] It should be noted that the process by which the electronic device obtains the second anomaly correction score based on the above-mentioned second comprehensive anomaly score, the above-mentioned benchmark anomaly score, and the above-mentioned preset anomaly correction coefficient can refer to the relevant description in step A6 above. To avoid repetition, this embodiment will not repeat it here.

[0104] Step 102d5: The electronic device uses the second anomaly correction score to correct the weighted duration and obtain the service duration of the first service.

[0105] It should be noted that the process of the electronic device using the above-mentioned second anomaly correction score to correct the weighted duration and obtain the service duration of the first service can be referred to the relevant description in step 102c2 above. To avoid repetition, this embodiment will not repeat it here.

[0106] In some embodiments of this disclosure, the data processing method provided in this disclosure may further include the following step 105: Step 105: If the first comprehensive abnormal score or the second comprehensive abnormal score is greater than the first preset score and less than or equal to the second preset score, adjust the duration of the first business based on the ratio between the business amount of the first business and the average business amount of at least one historical business and the preset amount correction coefficient.

[0107] In some embodiments of this disclosure, if the first comprehensive anomaly score or the second comprehensive anomaly score is greater than the first preset score and less than or equal to the second preset score, it indicates that the first comprehensive anomaly score or the second comprehensive anomaly score is relatively large. Therefore, the confidence level of the predicted service duration of the first service is not high enough. Consequently, the electronic device can use the preset amount sensitivity coefficient, the service amount of the first service, and the average service amount of at least one historical service to correct the service duration of the first service.

[0108] In some embodiments of this disclosure, the electronic device may use the following formula to correct the service duration of the first service described above: (15) in, This indicates the revised duration of the first service described above. This represents the preset monetary sensitivity coefficient. This represents the average transaction amount of at least one of the aforementioned historical transactions.

[0109] In some embodiments of this disclosure, the data processing method provided in this embodiment may further include the following step 106: Step 106: If the electronic device meets the first condition, it displays a prompt message. The prompt message is used to indicate that there is an abnormality in the first service. The first condition includes at least one of the following: The aforementioned first comprehensive anomaly score or the aforementioned second comprehensive anomaly score is greater than or equal to the aforementioned second preset score; The average of N duration deviations is greater than or equal to a first preset threshold. The aforementioned N duration deviations include the deviation between the predicted duration of each historical service in N consecutive historical services and the actual duration of that historical service. The attribute information of the aforementioned second entity indicates that the aforementioned second entity poses a risk; The aforementioned data on the first business indicates that the first business involves risks.

[0110] In some embodiments of this disclosure, if the first comprehensive abnormal score or the second comprehensive abnormal score is greater than or equal to the second preset score, it indicates that the first comprehensive abnormal score or the second comprehensive abnormal score is too high. Therefore, the first service may be abnormal, and the electronic device can display the first prompt information to prompt the user to process the first service.

[0111] In some embodiments of this disclosure, if the average of N duration deviations is greater than or equal to a first preset threshold, it indicates that the average of the service duration predicted in N consecutive historical data is too large, meaning that the service duration predicted in N consecutive historical data is not accurate enough. Therefore, there is an anomaly in the N consecutive historical data of services between the first entity and the second entity, and consequently, there may also be an anomaly in the first service between the first entity and the second entity. Consequently, the electronic device can display the first prompt information to prompt the user to process the first service.

[0112] In some embodiments of this disclosure, since the attribute information of the second subject indicates that the second subject is at risk, the first task between the first subject and the second subject may also be at risk. As a result, the electronic device can display the first prompt information to prompt the user to process the first service.

[0113] In some embodiments of this disclosure, since the first service data indicates that the first service is at risk, the electronic device can display the first prompt information to prompt the user to process the first service.

[0114] Step 103: The electronic device predicts the end time of the first service based on the service start time and service duration.

[0115] In some embodiments of this disclosure, the electronic device can extend the service start time by extending the service duration to obtain the end time of the first service.

[0116] For example, if the first business starts on August 1, 2026, and the electronic device predicts that the first business will last for 55 days, then the first business will end on September 25.

[0117] Step 104: The electronic device performs business risk prediction for the first entity based on the end time of all unfinished business between the first entity and all entities with which it has business relationships.

[0118] In some embodiments of this disclosure, prior to step 104 above, the data processing method provided in this disclosure may further include steps 107 and 108, and step 104 above can be implemented through step 104a below: Step 107: The electronic device determines the service duration interval in which the service duration of the first service falls based on the service duration of the first service, the comprehensive anomaly score, at least one actual duration of at least one historical service, and at least one predicted duration of at least one historical service.

[0119] In some embodiments of this disclosure, the aforementioned comprehensive anomaly score may be either the aforementioned first comprehensive anomaly score or the aforementioned second comprehensive anomaly score.

[0120] In some embodiments of this disclosure, the electronic device can determine the standard deviation of at least one predicted duration based on at least one actual duration and at least one predicted duration of at least one historical service, and then determine the service duration interval based on the standard deviation of the at least one predicted duration, the service duration, and the comprehensive anomaly score.

[0121] In some embodiments of this disclosure, the electronic device may use the following formula to calculate the standard deviation of at least one predicted duration: (16) in, The standard deviation represents the duration of at least one of the above predictions.

[0122] In some embodiments of this disclosure, the electronic device may use the following formula to determine the duration range of the aforementioned service: (17) in, This indicates the upper limit of the duration range of the aforementioned business. This represents the lower limit of the above-mentioned business duration interval, which can be understood as the confidence interval of the duration of the first business mentioned above. 1.96 represents the standard normal distribution quantile corresponding to the 95% confidence level.

[0123] Step 108: The electronic device determines the time interval of the end time based on the service duration interval and the service start time.

[0124] In some embodiments of this disclosure, the electronic device can postpone the start time of the service by the upper limit of the service duration interval to obtain the upper limit of the time interval in which the end time is located, and postpone the start time of the service by the lower limit of the service duration interval to obtain the lower limit of the time interval in which the end time is located.

[0125] Step 104a: The electronic device performs business risk prediction for the first entity based on the end time of all unfinished business between the first entity and all entities with which it has a business relationship, the time interval in which the end time of each unfinished business is located, the type of each unfinished business, and the entity type of the entity corresponding to each unfinished business.

[0126] In some embodiments of this disclosure, the electronic device can predict the balance of the first entity in the future based on the end time of all unfinished transactions between the first entity and all entities with which it has a business relationship, and the transaction amount of all unfinished transactions. It can also predict the balance range of the first entity in the future based on the time interval of the end time of each unfinished transaction in all unfinished transactions between the first entity and all entities with which it has a business relationship, and the transaction amount of all unfinished transactions.

[0127] In some embodiments of this disclosure, after the electronic device predicts the balance of the first entity and the balance range of the first entity in the future period, it can predict whether the balance of the first entity will be abnormal in the future period, and whether the balance range of the first entity will be abnormal in the future period, based on the type of each unfinished business in all unfinished businesses and the entity type of the entity corresponding to each unfinished business in all unfinished businesses.

[0128] In some embodiments of this disclosure, the electronic device may also obtain the average of the comprehensive abnormal scores of all unfinished transactions between the first entity and all entities with which it has a business relationship, and obtain the balance of the first entity at the current time. Then, based on the average of the M expenditure amounts of the first entity in M ​​consecutive time intervals of a preset historical duration, the average of the comprehensive abnormal scores of all unfinished transactions, and the balance of the first entity at the current time, the risk warning index of the first entity is determined.

[0129] In some embodiments of this disclosure, the electronic device may use the following formula to calculate the aforementioned risk warning index: (18) in, The above risk warning index indicates that... This indicates the current balance of the first entity. This represents the average of the M expenditure amounts mentioned above, and κ represents the risk amplification factor, with a default value of 0.3. This represents the average of the overall anomaly scores for all incomplete transactions.

[0130] In some embodiments of this disclosure, the aforementioned first service data also includes the agreed-upon end time of the first service. If the agreed-upon end time is earlier than the current time, the electronic device can determine that the aforementioned first service is overdue.

[0131] In some embodiments of this disclosure, the electronic device can identify all overdue transactions of all unfinished transactions between the first entity and all entities with which it has a business relationship, determine the overdue duration of each overdue transaction, and then determine the processing priority score of the overdue transaction based on the overdue duration, transaction amount, entity type of the entity corresponding to the overdue transaction, and the average actual duration of at least one historical transaction between the first entity and the entity corresponding to the overdue transaction.

[0132] In some embodiments of this disclosure, the electronic device may use the following formula to determine the above-mentioned processing priority score: (19) in, This represents the processing priority score, with a value ranging from 0 to 100; This indicates the aforementioned overdue period; A This indicates the amount of the overdue transaction; I The type score indicates the subject type of the overdue business. If the subject type of the overdue business indicates that there is a business level higher than the preset level between the overdue business subject and the first subject mentioned above, that is, the subject type of the overdue business subject is a high-level business subject type, the type score is 1; otherwise, the type score is 0. This indicates the average actual duration mentioned above. Dmax This represents the maximum number of overdue days, used for normalization, with a default value of 90. Amax This represents the maximum overdue amount, used for normalization, with a default value of 1,000,000. Hmax This represents the maximum duration of the service, used for normalization, with a default value of 90. , , , These represent the weights for overdue duration, overdue transaction amount, type score, and average actual duration, respectively, with default values ​​of 0.4, 0.4, 0.1, and 0.1.

[0133] In some embodiments of this disclosure, the electronic device can also perform parameter optimization periodically, as can be seen from the following formula: (20) in, This represents the optimized parameters. This represents the parameters before optimization. The learning rate has a default value of 0.05.

[0134] In some embodiments of this disclosure, the parameters for optimizing electronic devices may include industry-benchmark service duration. B New subject adjustment coefficient and prepayment adjustment coefficient The specifics can be determined based on actual needs, and this embodiment does not impose specific limitations here.

[0135] In the data processing method provided in this embodiment, a business dataset of a first entity is obtained. The business dataset includes first business data of an unfinished first business between the first entity and a second entity, the business quantity of at least one historical business between the first entity and the second entity, and at least one set of historical business data of the at least one historical business. The second entity is any entity that has a business relationship with the first entity, and the first business is any unfinished business between the first entity and the second entity. Based on the business quantity and at least one item in the at least one set of historical business data, and the business start time in the first business data, the duration of the first business is predicted. Based on the business start time and the duration of the business, the end time of the first business is predicted. Based on the end time of all unfinished businesses between the first entity and all entities with which it has a business relationship, a business risk prediction is performed on the first entity. Thus, the data processing method provided in this disclosure can predict the end time of an incomplete business by using any business data of the first entity's unfinished business in the first entity's business dataset, as well as historical business data between the first entity and the second entity corresponding to the unfinished business. After the electronic device predicts the end time of all unfinished businesses between the first entity and all entities with which it has business relationships, it can perform business risk prediction for the first entity based on the end time of all unfinished businesses between the first entity and all entities with which it has business relationships. As a result, the electronic device only needs to perform calculations on a small amount of data in the first entity's business dataset to perform business risk prediction for the first entity, thereby effectively saving computing resources and improving the efficiency of risk prediction.

[0136] The data processing method provided in this disclosure will now be described in conjunction with specific implementation methods.

[0137] Implementation method one: such as Figure 2 As shown, this embodiment provides a data processing system that can execute the above-described data processing method. The system adopts a three-layer main architecture, including a three-layer main structure consisting of a data layer 100, an intelligent layer 200, and an application layer 300, as well as two-sided support structures consisting of an import mechanism 400 and a feedback mechanism 500, and embeds an exception handling module 210 in the intelligent layer 200.

[0138] The data layer 100 includes a data acquisition module 110, an expenditure plan data module 120, and an external signal module 130. The data acquisition module 110 stores detailed information about the first entity's initiated business, including fields such as business number, entity name, entity type, business amount, business start date, business type, whether there is a prepaid business amount, and the prepaid amount percentage. This data is periodically imported from an Excel template or the enterprise ERP system via the import mechanism 400. The expenditure plan data module 120 stores the first entity's future expenditure arrangements, including fields such as expenditure item, expenditure amount, expenditure type, expenditure frequency, and expected expenditure date. This data is also maintained via the import mechanism 400. The external signal module 130 accesses external data such as industry prosperity indices, policy change information, and entity comment information. This data can be obtained through API interfaces or manual entry, providing data support for handling environmental anomalies.

[0139] The import mechanism 400 provides standardized data import interfaces, including an Excel import unit 410, a manual data entry unit 420, and a system integration unit 430. The Excel import unit 410 uses a predefined template structure; users only need to organize the data according to the template format to complete batch import. The system automatically performs data validation and cleaning, providing clear prompts for format errors or missing fields. The manual data entry unit 420 provides a form-style entry interface, supporting real-time addition and modification of small amounts of data, suitable for urgent business or temporary expenditure scenarios. The system integration unit 430 supports lightweight integration with mainstream ERP software, achieving automatic data synchronization via API calls or direct database connections.

[0140] The intelligent layer 200 is the core processing unit of the system, including a dual-mode prediction module 220, an anomaly handling module 210, and an action suggestion module 230. The dual-mode prediction module 220 consists of a fast-mode prediction unit 221, a slow-mode prediction unit 222, and a hybrid prediction mode unit 223 with anomaly correction. The fast-mode prediction unit 221 uses rule engine technology to quickly predict new entities and data-deficient scenarios based on industry benchmark business duration and business rules. The slow-mode prediction unit 222 uses a time-weighted average algorithm to accurately predict entities that are not new entities based on historical business data between the first entity and the new entity. The hybrid prediction mode unit 223 with anomaly correction dynamically adjusts the fusion weight of the fast and slow modes according to the historical business volume of a certain entity and integrates anomaly correction scores. Output the final prediction result.

[0141] The anomaly handling module 210 includes an anomaly detection unit 211 and an anomaly response unit 212. The anomaly detection unit 211 monitors various anomalies in real time, including periodic anomaly detection, monetary anomaly detection, business anomaly detection, and environmental anomaly detection. It uses Formula 14 and business rules to identify anomalies in external signals and calculates a comprehensive anomaly index. The anomaly response unit 212 automatically triggers data repair strategies, model switching strategies, or manual intervention processes based on the comprehensive anomaly index, ensuring that the system can still provide reliable prediction results under abnormal conditions.

[0142] The action suggestion module 230 includes a priority calculation unit 231 and an action list generation unit 232. The priority calculation unit 231 prioritizes overdue business based on a multi-factor weighted formula; the action list generation unit 232 automatically generates a daily to-do list based on the warning level and processing priority, and pushes it to relevant personnel.

[0143] The application layer 300 is the system's output presentation unit, comprising a business amount prediction module 310, a business amount water level chart module 320, and an action list module 330, which together constitute the "management dashboard." The business amount prediction module 310 displays monthly prediction data for the next 6-12 months in tabular form, and marks abnormal prediction data with anomaly markers and confidence intervals; the business amount water level chart module 320 visualizes the future business balance change trend in line chart form, and automatically marks three warning zones based on a dynamic risk warning index; the action list module 330 displays daily to-do tasks in list form, sorted by priority, with specific and clear task content.

[0144] The feedback mechanism 500 includes a deviation analysis unit 510 and a parameter optimization unit 520. The deviation analysis unit 510 periodically obtains the actual business duration from the application layer, compares it with historical prediction results, and calculates the prediction deviation trend for each entity and each time window. Based on the deviation analysis results, the parameter optimization unit 520 automatically adjusts model parameters such as the industry benchmark business duration and the adjustment coefficient for new entities, enabling continuous learning and self-optimization of the system.

[0145] In this embodiment, the connection relationship between the modules is as follows: the data layer 100 is connected to the external data source through the import mechanism 400, and the output end of the data layer 100 is connected to the input end of the intelligent layer 200; the output end of the intelligent layer 200 is connected to the input end of the application layer 300; the input end of the feedback mechanism 500 is connected to the actual result feedback interface of the application layer 300, and the output end of the feedback mechanism 500 is connected to the parameter adjustment interface of the intelligent layer 200.

[0146] Implementation Method 2: Specific Implementation of the Dual-Mode Prediction Module

[0147] For example, such as Figure 3 As shown in the figure, the workflow and specific implementation method of the dual-mode prediction module 220 are described in detail in this embodiment.

[0148] The input to the dual-mode prediction module 220 comes from the data acquisition module 110 of the data layer 100. For each business to be predicted, historical business data is first queried. Specifically, the number of past businesses n and the duration of past businesses of the entity corresponding to the business are counted from the historical payment record table. If n is 0 or less than the minimum number of learning samples Nmin (Nmin is 3 in this embodiment), the fast mode prediction path is entered; if n is greater than or equal to the maximum number of learning samples Nmax (Nmax is 10 in this embodiment), the slow mode prediction path is entered; if n is between the two, the hybrid prediction path with anomaly correction is entered.

[0149] In the fast-mode prediction path, a rule engine prediction formula is used for calculation. In this embodiment, the industry benchmark business duration is... Based on the industry of the first entity, let's assume it's manufacturing. The value is 45 days; the second subject type is "new subject", therefore The value is 1.2; there is no prepayment for the business, therefore... The value is 1.0; the transaction amount is 300,000 yuan, which is greater than the small amount threshold of 100,000 yuan, therefore... The value is 1.0; the industry characteristic coefficient f is 1.0. Substituting into formula (1), we get: the duration of the first business above = 45 × 1.2 × 1.0 × 1.0 × 1.0 = 54 days.

[0150] In the slow-mode prediction path, the prediction formula of the learning module is used for calculation. Taking an entity that is not a new topic, i.e., an old topic, as an example, its historical repayment records are shown in Table 1: Table 1. Historical Business Duration Records of a Certain Entity

[0151] The current date is February 1, 2025, and the new business amount M is 250,000 yuan. First, calculate the time weight of each historical business, taking λ=0.01: Transaction 1: The difference in days is 22 days, w1 = = = 0.8025 Transaction 2: Day difference 88 days, w2 = = = 0.4148 Transaction 3: Day difference 165 days, w3 = = = 0.1920 Transaction 4: Day difference 262 days, w4 = = = 0.0728 The weights sum ∑wi = 0.8025 + 0.4148 + 0.1920 + 0.0728 = 1.4821 Calculate the weighted average ratio: ∑(wi×ri) = 0.8025×4878 + 0.4148×3750 + 0.1920×6522 + 0.0728×2778 = 3915 + 1556 + 1252 + 202 = 6925 = 6,925 / 1.4821 = 4672 yuan / day Calculate the trend correction factor by taking the three most recent transactions (transactions 1-3): The actual cycle for Transaction 1 is 41 days, the theoretical cycle is 200,000 / 4672 = 42.8 days, and the deviation rate is -4.2%. Business 2 has an actual cycle of 40 days, a theoretical cycle of 150,000 / 4,672 = 32.1 days, and a deviation rate of +24.6%. The actual cycle for business item 3 was 46 days, the theoretical cycle was 300,000 / 4672 = 64.2 days, and the deviation rate was -28.3%. The average deviation rate δ trend = (-0.042 + 0.246 - 0.283) / 3 = -0.079 / 3 = -0.0263 Computational learning predicts the duration of the second business. : = 250,000 / 4,672 × (1 - 0.0263) = 53.5 × 0.9737 = 52.1 days ≈ 52 days In the hybrid prediction path with anomaly correction, hybrid prediction is performed, specifically calculated using a hybrid prediction formula with anomaly correction. Assuming a customer has 8 historical transactions (n=8), the static learning weights... = (8-3) / (10-3) = 5 / 7 ≈ 0.714. (Rule prediction) =48 days, learning prediction =52 days. Simultaneously, industry anomaly scores are obtained from the anomaly handling module. =20 (indicating a normal industry environment) and comprehensive abnormal score =25 (indicating that the business itself has no obvious abnormalities, lower than the benchmark value of 30). The dynamic learning weight (i.e., the second weight mentioned above) and the anomaly correction score are calculated according to the formula. : = × (1 - η · = 0.714 × (1 - 0.5×20 / 100) = 0.643 = 1 + 0.2×(25-30) / 100 = 0.99 Substitute into the mixed prediction formula: = [ ] × γ abnormality = [0.643×52 + (1-0.643)×48] × 0.99 ≈ 50 days The output of the dual-mode prediction module 220 is passed to the anomaly handling module 210 for anomaly handling, and also to the action suggestion module 230 for subsequent processing. This embodiment, through the coordinated operation of fast and slow modes, ensures rapid response to new subject scenarios while achieving accurate prediction of non-new subject scenarios. By using hybrid prediction to achieve a smooth transition, it effectively solves the limitations of a single prediction model.

[0152] Implementation Method 3: Computational Analysis of Hybrid Prediction with Anomaly Correction

[0153] This embodiment details the specific calculation process of the hybrid prediction formula with anomaly correction, and demonstrates its technical effectiveness through comparative analysis.

[0154] Taking the second entity, "Entity C," as an example, this entity has 8 historical business transactions, falling within the mixed forecast range. Its current new business amount is 1.2 million yuan, and its historical average business amount is 400,000 yuan. Fast-mode forecasting. =48 days, slow mode prediction =60 days. = =(8-3) / (10-3)=5 / 7≈0.714.

[0155] Scenario 1: Normal environment, normal business operations

[0156] Assuming industry anomaly score =20 (Contract environment is favorable), Overall abnormal score =25 (lower than the baseline of 30). Calculate the dynamic learning weights: =0.714×(1-0.5×20 / 100)=0.714×0.9=0.643 Calculate the anomaly correction factor: =1 + 0.2 × (25 - 30) / 100 = 1 + 0.2 × (-0.05) = 1 - 0.01 = 0.99 Calculate the mixed prediction results: =[0.643×60+(1-0.643)×48]×0.99=(38.58+17.14)×0.99=55.72×0.99=55.16 days In contrast, the results of traditional hybrid prediction (without considering anomalies) are as follows: =0.714×60+0.286×48=42.84+13.73=56.57 days The difference is only 1.41 days, which is not significant because the environment and business are normal.

[0157] Scenario 2: Environmental anomaly, business operations normal

[0158] Assuming the industry is in a downturn, the industry's abnormal score is... =80; The business itself is normal. =25.

[0159] Calculate the dynamic learning weights: =0.714×(1-0.5×80 / 100)=0.714×0.6=0.428 The anomaly correction factor remains unchanged. =0.99.

[0160] Calculate mixed predictions: =[0.428×60+(1-0.428)×48]×0.99=(25.68+27.46)×0.99=53.14×0.99=52.61 days Traditional hybrid forecasting remains at 56.57 days. This implementation's forecast is approximately 4 days shorter than the traditional forecast because the system reduces its reliance on historical data during periods of industry downturn (learning weights decrease from 0.714 to 0.428), relying more on rule-based forecasting (48 days), making the forecast more conservative. Actual tracking shows that during periods of industry downturn, the actual duration of this historically recorded business was 53 days, with the forecast error of this invention being only 0.39 days, while the traditional forecast error reached 3.57 days.

[0161] Scenario 3: Normal environment, abnormal business operation

[0162] Assuming a normal industry environment 20, but the business amount is unusually high (1.2 million, far exceeding the historical average of 400,000), resulting in an overall abnormal score. =65 (Above 60 triggers manual intervention, but can still be automatically corrected within the range of cautious use).

[0163] Calculate the dynamic learning weights: =0.643 (Same as Scenario 1)

[0164] Calculate the anomaly correction factor: =1 + 0.2 × (65 - 30) / 100 = 1 + 0.2 × 0.35 = 1 + 0.07 = 1.07 Calculate mixed predictions: T =[0.643×60+0.357×48]×1.07=(38.58+17.14)×1.07=55.72×1.07=59.62 days The traditional hybrid forecast remains at 56.57 days. The forecast result of this implementation is approximately 3 days longer than the traditional forecast due to the monetary anomaly factor. =1.07 amplifies the prediction result by 7%, reflecting the business pattern that "large transactions last longer." Actual tracking shows that the actual duration of this transaction was even longer, at 61 days. The prediction error of this invention was only 1.38 days, while the traditional prediction error reached 4.43 days.

[0165] Scenario 4: Abnormal industry environment and business operations

[0166] Assuming an abnormal environment =80, business error. =65.

[0167] Calculate the dynamic learning weights: =0.428 (Same as Scenario 2)

[0168] Calculate the anomaly correction factor: =1.07 (Same as Scenario 3)

[0169] Calculate mixed predictions: T =[0.428×60+0.572×48]×1.07=(25.68+27.46)×1.07=53.14×1.07=56.86 days Traditional hybrid forecasts have a duration of 56.57 days, which is close to the traditional forecast. However, the forecasting mechanism of this implementation method is completely different: traditional forecasts rely on historical data (weight 0.714), while this implementation method reduces the historical weight (0.428) when the industry environment is abnormal, and at the same time amplifies the results through an anomaly correction factor (1.07). The combined effect of the two makes the result coincidentally similar to that of traditional forecasts, but the logic is more in line with reality—it takes into account both the need for conservatism due to environmental anomalies and the need for extension due to abnormal monetary amounts.

[0170] Implementation Method 4: Specific Implementation of the Application Layer "Management Dashboard"

[0171] like Figure 5 As shown in the figure, this embodiment describes in detail the business amount water level chart of the "Management Dashboard" in the application layer 300.

[0172] The business amount forecast module 310 displays the monthly forecast data for the next 6 months in tabular form. The lower and upper limits of the ending balance are calculated by summing the lower and upper limits of the confidence interval for the duration of the business forecast for each month. The confidence interval is calculated according to formula (17). As shown in Table 2: Table 2 Business Amount Forecast Table (including anomaly markers)

[0173] In March, the "Attention" marker indicates that some of the business on which the forecast relied for that month has a moderate anomaly (such as a large transaction by a certain entity for the first time), and the system suggests using the forecast data for that month with caution. In May, the "Anomaly" marker indicates that a certain business on which the forecast relied for that month has a serious anomaly (such as the entity corresponding to the business having overdue payment risks), and the system highlights it with the text "Anomaly" and adds the note: "Source of anomaly: High risk of overdue payment for Jia's drawing business, verification recommended." Each anomaly marker can be clicked to view details, including specific information about the abnormal business, the type of anomaly, and corresponding response suggestions.

[0174] The Business Amount Water Level Module 320 displays future business amount trends in a line chart format, with the horizontal axis representing months and the vertical axis representing business amount (ten thousand yuan). Three lines are plotted in the chart: the forecast line (solid line) represents the most likely scenario, the upper limit line (dashed line) represents the optimistic scenario (based on the forecast upper limit), and the lower limit line (dotted line) represents the pessimistic scenario (based on the forecast lower limit). Simultaneously, a warning line is plotted at 200,000 yuan on the vertical axis, representing the minimum safe business amount reserve. The safety line is defined at 500,000 yuan on the vertical axis. The chart automatically divides the area into three zones based on the risk warning index: low-risk zone, medium-risk zone, and high-risk zone. The low-risk zone corresponds to the portion where the forecast line is more than 3 times higher than the safety line; the medium-risk zone corresponds to the portion where the forecast line is between 1.5 and 3 times higher than the safety line; and the high-risk zone corresponds to the portion where the forecast line is less than 1.5 times lower than the safety line. The system adds "Low Risk," "Medium Risk," and "High Risk" labels to the edges of the corresponding zones and clearly explains them in the legend. In this example, the business balance in February was 1.6 million, which is in the low-risk zone; the business balance in March and April was 1.1-1.45 million, which is in the medium-risk zone; and the balance after May was 1.05-1.2 million, which is also in the medium-risk zone. Overall, the situation is under monitoring, and no high-risk areas have emerged.

[0175] Action List module 330 displays daily tasks in list format, as shown in Table 3: Table 3 Today's Action List

[0176] The action list is automatically generated at 9:00 AM daily and dynamically updated based on real-time data changes. Completed tasks can be marked as "Completed," and the system records the completion status for subsequent performance analysis. If a red alert is issued, an urgent task will be automatically inserted into the action list, such as "[Urgent] Initiate financing channels and contact XX Bank to apply for credit."

[0177] This embodiment transforms complex business transaction data and risk information into intuitive charts and actionable instructions through the design of a "management dashboard," greatly improving processing efficiency.

[0178] Example 5: Parameter Optimization Process of Feedback Mechanism

[0179] like Figure 6 As shown in the figure, this embodiment describes in detail the parameter optimization process of the feedback mechanism 500.

[0180] The parameter optimization process of feedback mechanism 500 is executed once a month, and the specific steps are as follows: Step S601: Collect the actual business data completed in the previous month.

[0181] For example, from the data acquisition module 110 of the data layer 100, the business transactions completed in the previous month are filtered out, and the actual duration of each business transaction is obtained. .

[0182] Step S602: Match historical prediction data.

[0183] For example, query the system logs for the predicted duration of these services when they were predicted last month. This forms a prediction-actual comparison pair.

[0184] Step S603: Calculate the deviation trend of each parameter.

[0185] For example, regarding the industry benchmark business duration Analyze the relative deviations of all business operations:

[0186] Where N is the total number of transactions completed in the current month.

[0187] In this embodiment, a total of 20 transactions were completed last month, and the average deviation was calculated. = +0.052, indicating that the overall forecast is 5.2% more optimistic.

[0188] Step S604: Update model parameters.

[0189] For example, a learning rate can be used. η =0.05, calculate the new industry benchmark payment period: New = old(1+ η × =45(1 + 0.05 × 0.052) = 45(1 + 0.0026) = 45.12 days ≈ 45 days (rounded down) For the time decay coefficient λ, we analyze the difference between the deviations of recent and long-term business. The average deviation δrecent of the business over the most recent 3 months is calculated as +3.2%, and the average deviation δlong-term of the business over the previous 3 months is +6.8%. Since δrecent < δlong-term, it indicates that the recent deviation is narrowing and business is accelerating. Therefore, the weight of recent data should be appropriately increased; hence, λ is adjusted from 0.01 to 0.0105 (an increase of 5%).

[0190] For the new subject coefficient The analysis shows that the average deviation of the new main business is δ = +8.5%, indicating that the existing coefficient of 1.2 is still too optimistic. Therefore, it is adjusted to: = 1.2 × (1 + 0.05 × 0.085) = 1.2 × 1.00425 = 1.205 Step S605: Update the rule coefficients.

[0191] For example, other rule coefficients such as prepayment coefficients and small-amount acceleration coefficients can also be adjusted in a similar way, but these coefficients are usually relatively stable and do not need to be adjusted frequently.

[0192] Step S606: Record the optimization results.

[0193] For example, the electronic device can save the updated parameters to the parameter configuration table and record the deviation analysis and parameter adjustment log of this optimization for subsequent auditing and backtracking.

[0194] This embodiment enables the system to adaptively track changes in the business environment through periodic optimization of the feedback mechanism, and the prediction accuracy continues to improve with the extension of usage time.

[0195] Implementation Method Six: Overall System Workflow

[0196] like Figure 7 As shown, this embodiment illustrates the overall workflow of the system through a complete timing diagram.

[0197] At 9:00 AM on Monday, the system automatically starts the weekly business forecasting process. First, the import mechanism 400 reads the latest Excel data files from the specified folder, including the business order table, historical payment record table, and expenditure plan table, performs data format verification and cleaning, and loads the data into the corresponding modules of the data layer 100.

[0198] After the data loading is complete, the intelligent layer 200 starts the dual-mode prediction module 220, which iterates through all unfinished business processes, sequentially performing customer historical data queries, fast / slow mode selection, and prediction calculations. For each business process, the prediction result, along with the original business data, is passed to the anomaly handling module 210 for anomaly detection, calculation of the comprehensive anomaly index, and determination of response strategies. For business processes requiring manual intervention, the system pauses automatic processing and generates an audit task at the application layer 300; for other business processes, the system uses the corrected prediction results for subsequent processing. Simultaneously, the hybrid prediction unit 223 uses the calculated comprehensive anomaly index to generate an anomaly correction factor. It is then directly applied to the prediction formula to achieve anomaly detection prediction.

[0199] After all business transactions are completed, the Action Recommendation module 230 summarizes the predicted end dates for each transaction and groups them by month to calculate the predicted transaction amounts. Simultaneously, it retrieves monthly expenditure forecasts from the Expenditure Planning Data module 120, calculates the monthly transaction amount and the end-of-period transaction amount, and generates a transaction amount forecast table. The Risk Warning module calculates a risk index based on the end-of-period transaction amount and the average monthly net outflow to determine the warning level. The Priority module analyzes overdue transactions, calculates priority scores, and ranks them.

[0200] The application layer 300 receives the output from the intelligence layer 200, renders and generates a business amount forecast table, a business amount water level chart, and an action list, and pushes the warning information and action list to designated personnel. The boss can open his mobile phone to view today's "management dashboard" to understand the company's business status and to-do items.

[0201] During system operation, if new business agreements are signed or expenditure plans are adjusted, users can update data at any time through the import mechanism 400. The system will automatically trigger incremental calculations to update forecast results and early warning status.

[0202] Every Friday at 5:00 PM, the system automatically generates a weekly report, summarizing the comparison between the week's business forecast and actual results, the handling of abnormal business, the completion status of collection tasks, etc., and sends it to relevant personnel.

[0203] At the beginning of each month, the feedback mechanism 500 automatically executes the parameter optimization process, analyzes the prediction deviation of the previous month, updates the model parameters, and records the optimization results in the system log.

[0204] The above workflow fully covers the entire closed-loop process from data import, predictive calculation, anomaly handling, result display to model optimization. Detailed steps are as follows: Step 701: The user uploads an Excel file to the import mechanism module, triggering the system to start processing the data.

[0205] Step 702: The import mechanism sends a data validation request to the data layer to check whether the Excel format and content meet the requirements. The data layer then returns the validation results to the import mechanism.

[0206] Step 703: After the verification is passed, the import mechanism will officially load the data into the data layer and store it in the corresponding business order, expenditure plan and other modules.

[0207] Step 704: The data layer sends a read request to the intelligence layer, and the intelligence layer returns the business data. This step will process each unfinished business in a loop.

[0208] Step 705: For each service, the data layer sends a prediction request to the intelligence layer, requesting the calculation of the expected duration of the service.

[0209] Step 706: The intelligent layer requests the comprehensive anomaly index from the anomaly handling layer for hybrid prediction correction (fast mode, slow mode, or hybrid mode). The anomaly handling layer returns the prediction result (Note: In the diagram, the return of the prediction result is represented as anomaly handling layer → intelligent layer).

[0210] Step 707: The intelligent layer initiates an exception handling request to the exception handling department, requesting a multi-angle exception analysis of the current business.

[0211] Step 708: Calculate the comprehensive anomaly score (for anomaly handling) EI The score is based on four aspects: cycle, amount, business, and environment.

[0212] Step 709, Exception Handling EI The value determines the response strategy (trust model, cautious use, or manual intervention).

[0213] Step 710: Anomaly handling returns the anomaly detection results (including correction suggestions and anomaly correction factors) to the intelligent layer.

[0214] Step 711: The intelligent layer stores the final prediction results of each business to the application layer for subsequent display.

[0215] Step 712: The data layer requests the expenditure plan data from the intelligence layer, and the intelligence layer returns the expenditure data (for subsequent business amount calculation).

[0216] Step 713: The intelligent layer summarizes the forecast results of all services and compiles them monthly.

[0217] Step 714: The intelligent layer calculates the business amount (revenue minus expenses) for each future month.

[0218] Step 715: The intelligent layer calculates the dynamic risk warning index and determines the warning level.

[0219] Step 716: The intelligent layer calculates the priority score of overdue services and generates a processing list.

[0220] Step 717: The application layer pushes the generated prediction report and action list to the feedback mechanism as input for subsequent optimization.

[0221] Step 718: Users view the "Management Dashboard" (Business Amount Forecast Table, Business Amount Level Chart, Action List) displayed in the application layer to obtain decision-making information.

[0222] Step 719: The feedback mechanism automatically executes the parameter optimization process monthly, adjusting model parameters based on the actual duration of business operations (such as the industry benchmark business duration) to achieve system self-learning.

[0223] This embodiment achieves a complete closed loop from data input to decision output through a fully automated processing flow, minimizing manual intervention and improving work efficiency. The system only requires relevant personnel to spend a short time updating Excel data each week to automatically generate all forecast and early warning information, greatly improving forecasting efficiency.

[0224] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art can make various improvements and modifications without departing from the spirit and principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

[0225] Figure 8 This is a block diagram illustrating a data processing apparatus according to some embodiments of the present disclosure. (Refer to...) Figure 8 The data processing device 800 includes an acquisition module 801 and a prediction module 802.

[0226] The acquisition module 801 is used to acquire the business dataset of the first subject. The business dataset includes the first business data of the first business that has not been completed between the first subject and the second subject, the business quantity of at least one historical business between the first subject and the second subject, and at least one set of historical business data of the at least one historical business. The second subject is any subject that has a business relationship with the first subject, and the first business is any business that has not been completed between the first subject and the second subject. The prediction module 802 is used to predict the duration of the first service based on the number of services and at least one item in the at least one set of historical service data, as well as the service start time in the first service data. The prediction module 802 is further configured to predict the end time of the first service based on the service start time and the service duration. The prediction module 802 is further configured to predict business risks for the first entity based on the end time of all unfinished business between the first entity and all entities with which it has business relationships.

[0227] In some embodiments of this disclosure, the prediction module 802 is specifically used for: If the number of services is less than or equal to a first preset number, then based on the industry characteristic coefficient of the industry in which the first entity operates and the industry benchmark service duration of the industry in which the first entity operates, the duration of the first service is determined, and the duration of the first service is determined as the service duration of the first service; or, If the number of services is greater than or equal to the second preset number, the duration of the second service is determined based on the at least one set of historical service data. Based on the duration of the second service and the attribute information of the second entity, the duration of the first service is determined; or, If the number of services is greater than the first preset number and less than the second preset number, then the duration of the first service is determined based on the duration of the first service, the duration of the second service, and the attribute information. The second preset quantity is greater than the first preset quantity.

[0228] In some embodiments of this disclosure, each set of historical business data includes the start time, end time, and amount of the corresponding historical business; the prediction module 702 is specifically used for: The actual duration of each historical service is determined based on its start and end times. Calculate the first ratio between the transaction amount of each historical transaction and the actual duration of that historical transaction; Calculate the difference between the actual duration of each historical service and the predicted duration of that historical service, and the second ratio between the difference in duration of that historical service and the predicted duration of that historical service. The duration of the second service is determined based on the sum of at least one first ratio corresponding to the at least one historical service, the service amount of the first service, and at least one second ratio corresponding to the at least one historical service.

[0229] In some embodiments of this disclosure, the prediction module 802 is specifically used for: Based on the attribute information of the second subject, the service type of the first service, the at least one set of historical service data, and the duration of the second service, a first anomaly correction score is determined; The duration of the second service is corrected using the first anomaly correction score to obtain the duration of the first service.

[0230] In some embodiments of this disclosure, the attribute information includes the subject type of the second subject and the comment information of the second subject, and the prediction module 802 is specifically used for: Based on the subject type and the business type, determine the business anomaly score of the first business; Based on the statistical value of at least one actual duration of the at least one historical service and the duration of the second service, determine the duration anomaly score of the first service. Based on at least one transaction amount of the at least one historical transaction and the transaction amount of the first transaction, determine the abnormal transaction amount score of the first transaction; Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. The abnormal business score, the abnormal industry score, the abnormal duration score, and the abnormal business amount score are weighted and calculated to obtain a comprehensive abnormal score. The first abnormality correction score is obtained based on the comprehensive abnormality score, the benchmark abnormality score, and the preset abnormality correction coefficient.

[0231] In some embodiments of this disclosure, the device 800 further includes: a correction module 804, configured to: If the overall anomaly score is greater than the first preset score and less than or equal to the second preset score, the duration of the first service is adjusted based on the service amount of the first service, the average service amount of the at least one historical service, and a preset amount correction coefficient. The second preset score is greater than the first preset score.

[0232] In some embodiments of this disclosure, the device further includes: a display module 805, used for...

[0233] If the first condition is met, a prompt message is displayed, which is used to indicate that the first service is abnormal; The first condition includes at least one of the following: The overall anomaly score is greater than or equal to the first preset score; The average of N duration deviations is greater than or equal to the second preset threshold, wherein the N duration deviations include the deviation between the predicted duration of each historical service in N consecutive historical services and the actual duration of that historical service; The attribute information of the second subject indicates that the second subject is at risk; The first business data indicates that the first business is at risk.

[0234] In some embodiments of this disclosure, the prediction module 802 is further configured to: Before predicting business risks for the first entity, based on the end time of all unfinished business between the first entity and all entities with which it has a business relationship, the business duration interval of the first business is determined based on the business duration of the first business, the comprehensive anomaly score, at least one actual duration of the at least one historical business, and at least one predicted duration of the at least one historical business. Based on the duration range of the service and the start time of the service, the time range in which the end time falls is determined; The prediction module 802 is specifically used for: Based on the end time of all unfinished business between the first entity and all entities with which it has a business relationship, the time interval in which the end time of each unfinished business falls, the type of each unfinished business, and the entity type of the entity corresponding to each unfinished business, business risk prediction is performed on the first entity.

[0235] In some embodiments of this disclosure, the attribute information includes the comment information of the second subject and the subject type of the second subject, and the prediction module 702 is specifically used for: Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. Based on the industry anomaly score, a first weight value for the duration of the first business and a second weight value for the duration of the second business are determined. The first weight value and the second weight value are used to perform a weighted calculation on the duration of the first service and the duration of the second service to obtain a weighted duration. A second anomaly correction score is determined based on the industry anomaly score, the subject type, the business type, the at least one set of historical business data, and the weighted duration. The weighted duration is corrected using the second anomaly correction score to obtain the service duration of the first service.

[0236] Since the detailed implementation of each module in the data processing apparatus provided in the exemplary embodiments of this disclosure has been described in the exemplary embodiments of the data processing method described above, it will not be repeated here.

[0237] It should be noted that although several modules or units of the data processing apparatus have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0238] Exemplary embodiments of this disclosure also provide a computer program product. The computer program product includes a computer program that, when executed by a processor, implements the data processing method described above.

[0239] In one implementation, the computer program product can be a tangible product containing a computer program, such as a computer-readable storage medium storing the computer program. The readable storage medium can be a storage medium based on electrical, magnetic, optical, electromagnetic, infrared, or other signals, including but not limited to: random access memory (RAM), read-only memory (ROM), magnetic tape, floppy disk, flash memory, hard disk drive (HDD), solid-state drive (SSD), etc. For example, the computer program product can be implemented as a non-volatile storage medium storing a computer program, such as read-only memory, NAND flash memory, etc.

[0240] In one implementation, the computer program product can be an intangible product containing a computer program. For example, the computer program product can be implemented as a virtual digital product, such as an executable file, installation package, or other digital file storing the computer program.

[0241] Computer program code can be written in one or more programming languages. Examples of programming languages ​​include C, Java, and C++. Program code can execute entirely on the user's computing device, partially on the user's computing device, or as a standalone software package. It can also execute partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, such as a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via an internet connection provided by a mobile network operator).

[0242] Computer programs can be carried or transmitted via signals such as electrical, magnetic, optical, electromagnetic, and infrared rays. Electronic devices can convert signals carrying computer programs into digital signals, thereby running the computer programs. When a computer program runs on an electronic device, its code is used to cause the electronic device to execute (more specifically, the processor of the electronic device to execute) the method steps of various exemplary embodiments of this disclosure, such as the data processing method described above.

[0243] Furthermore, in an exemplary embodiment of this disclosure, an electronic device capable of implementing the above-described method is also provided, comprising: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to perform an exemplary method of any of the above-described methods by executing the executable instructions.

[0244] The following is for reference. Figure 9 The electronic device is illustrated by way of a general-purpose computing device. It should be understood that... Figure 9The electronic device 900 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0245] like Figure 9 As shown, the electronic device 900 may include: a processor 910, a memory 920, a bus 930, an I / O (input / output) interface 940, and a network adapter 950.

[0246] The memory 920 may include volatile memory, such as RAM 921 and cache unit 922, and may also include non-volatile memory, such as ROM 923. The memory 920 may also include one or more program modules 924, including but not limited to: an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. For example, program module 924 may include the modules described above.

[0247] The processor 910 may include one or more processing units, such as an AP (Application Processor), a modem processor, a GPU (Graphics Processing Unit), an ISP (Image Signal Processor), a controller, an encoder, a decoder, a DSP (Digital Signal Processor), a baseband processor, and / or an NPU (Neural-Network Processing Unit).

[0248] The processor 910 can be used to execute executable instructions stored in the memory 920 to perform method steps of various embodiments of this disclosure, such as... Figure 1 The steps are shown.

[0249] Bus 930 is used to connect different components of electronic device 900 and may include data bus, address bus and control bus.

[0250] Electronic device 900 can communicate with one or more external devices 1000 (such as keyboard, mouse, external controller, etc.) through I / O interface 940.

[0251] Electronic device 900 can communicate with one or more networks via network adapter 950. For example, network adapter 950 can provide mobile communication solutions such as 3G / 4G / 5G, or wireless communication solutions such as wireless LAN, Bluetooth, and near-field communication. Network adapter 950 can communicate with other modules of electronic device 900 via bus 930.

[0252] In one embodiment, the electronic device 900 further includes a display for displaying a graphical user interface.

[0253] although Figure 9 As not shown in the diagram, other hardware and / or software modules may also be configured in the electronic device 900, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID (Redundant Arrays of Independent Disks) systems, tape drives, and data backup storage systems.

[0254] As can be seen from the above, the technical solutions disclosed herein can be implemented as methods, apparatus, systems, computer program products, storage media, electronic devices, etc. Those skilled in the art will understand that various aspects of this disclosure can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or an implementation combining hardware and software aspects, which may be referred to as "circuit," "module," or "system," respectively.

[0255] It should be understood that this disclosure is not limited to the specific methods, steps, or structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. Those skilled in the art will readily conceive of other embodiments based on the specific implementations provided in this disclosure. Therefore, the specific implementations provided in this disclosure are merely exemplary, and the scope and spirit of this disclosure are indicated by the claims, and should cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary technical means in the art not disclosed in this disclosure.

Claims

1. A data processing method, characterized in that, include: Obtain the business dataset of the first entity, the business dataset including the first business data of the first business between the first entity and the second entity that has not been completed, the business quantity of at least one historical business between the first entity and the second entity, and at least one set of historical business data of the at least one historical business, the second entity being any entity that has a business relationship with the first entity, and the first business being any business between the first entity and the second entity that has not been completed; Based on the number of services and at least one item from the at least one set of historical service data, as well as the service start time in the first service data, the duration of the first service is predicted. Based on the start time and duration of the service, predict the end time of the first service; Based on the end time of all unfinished business between the first entity and all entities with which it has business relationships, business risk is predicted for the first entity.

2. The method according to claim 1, characterized in that, The step of predicting the duration of the first service based on the number of services and at least one item from the at least one set of historical service data, as well as the service start time in the first service data, includes: If the number of services is less than or equal to a first preset number, then based on the industry characteristic coefficient of the industry in which the first entity operates and the industry benchmark service duration of the industry in which the first entity operates, the duration of the first service is determined, and the duration of the first service is determined as the service duration of the first service; or, If the number of services is greater than or equal to the second preset number, the duration of the second service is determined based on the at least one set of historical service data. Based on the duration of the second service and the attribute information of the second entity, the duration of the first service is determined; or, If the number of services is greater than the first preset number and less than the second preset number, then the duration of the first service is determined based on the duration of the first service, the duration of the second service, and the attribute information. The second preset quantity is greater than the first preset quantity.

3. The method according to claim 2, characterized in that, Each set of historical business data includes the start time, end time, and amount of the corresponding historical business; determining the duration of the second business based on the at least one set of historical business data includes: The actual duration of each historical service is determined based on its start and end times. Calculate the first ratio between the transaction amount of each historical transaction and the actual duration of that historical transaction; Calculate the difference between the actual duration of each historical service and the predicted duration of that historical service, and the second ratio between the difference in duration of that historical service and the predicted duration of that historical service. The duration of the second service is determined based on the sum of at least one first ratio corresponding to the at least one historical service, the service amount of the first service, and at least one second ratio corresponding to the at least one historical service.

4. The method according to claim 2 or 3, characterized in that, Determining the duration of the first service based on the duration of the second service and the attribute information of the second subject includes: Based on the attribute information of the second subject, the service type of the first service, the at least one set of historical service data, and the duration of the second service, a first anomaly correction score is determined; The duration of the second service is corrected using the first anomaly correction score to obtain the duration of the first service.

5. The method according to claim 4, characterized in that, The attribute information includes the subject type of the second subject and the comment information of the second subject. Determining the first anomaly correction score based on the attribute information of the second subject, the business type of the first business, the at least one set of historical business data, and the duration of the second business includes: Based on the subject type and the business type, determine the business anomaly score of the first business; Based on the statistical value of at least one actual duration of the at least one historical service and the duration of the second service, determine the duration anomaly score of the first service. Based on at least one transaction amount of the at least one historical transaction and the transaction amount of the first transaction, determine the abnormal transaction amount score of the first transaction; Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. The abnormal business score, the abnormal industry score, the abnormal duration score, and the abnormal business amount score are weighted and calculated to obtain a comprehensive abnormal score. The first abnormality correction score is obtained based on the comprehensive abnormality score, the benchmark abnormality score, and the preset abnormality correction coefficient.

6. The method according to claim 5, characterized in that, The method further includes: If the overall anomaly score is greater than the first preset score and less than or equal to the second preset score, the duration of the first service is adjusted based on the service amount of the first service, the average service amount of the at least one historical service, and a preset amount correction coefficient. The second preset score is greater than the first preset score.

7. The method according to claim 6, characterized in that, The method further includes: If the first condition is met, a prompt message is displayed, which is used to indicate that the first service is abnormal; The first condition includes at least one of the following: The overall anomaly score is greater than or equal to the first preset score; The average of N duration deviations is greater than or equal to the second preset threshold, wherein the N duration deviations include the deviation between the predicted duration of each historical service in N consecutive historical services and the actual duration of that historical service; The attribute information of the second subject indicates that the second subject is at risk; The first business data indicates that the first business is at risk.

8. The method according to claim 5 or 6, characterized in that, Before performing business risk prediction for the first entity based on the end times of all outstanding business transactions between the first entity and all entities with which it has business relationships, the method further includes: Based on the service duration of the first service, the comprehensive anomaly score, at least one actual duration of the at least one historical service, and at least one predicted duration of the at least one historical service, the service duration interval in which the service duration of the first service falls is determined. Based on the duration range of the service and the start time of the service, the time range in which the end time falls is determined; The business risk prediction for the first entity based on the end times of all outstanding business transactions between the first entity and all entities with which it has business relationships includes: Based on the end time of all unfinished business between the first entity and all entities with which it has a business relationship, the time interval in which the end time of each unfinished business falls, the type of each unfinished business, and the entity type of the entity corresponding to each unfinished business, business risk prediction is performed on the first entity.

9. The method according to claim 2, characterized in that, The attribute information includes the comment information of the second subject and the subject type of the second subject. Determining the duration of the first service based on the duration of the first service, the duration of the second service, and the attribute information of the second subject includes: Based on the comment information, the industry prosperity index and industry risk index of the industry in which the second subject is located, the industry anomaly score of the industry in which the second subject is located is determined. Based on the industry anomaly score, a first weight value for the duration of the first business and a second weight value for the duration of the second business are determined. The first weight value and the second weight value are used to perform a weighted calculation on the duration of the first service and the duration of the second service to obtain a weighted duration. A second anomaly correction score is determined based on the industry anomaly score, the subject type, the business type, the at least one set of historical business data, and the weighted duration. The weighted duration is corrected using the second anomaly correction score to obtain the service duration of the first service.

10. A data processing apparatus, characterized in that, include: The acquisition module is used to acquire the business dataset of the first subject. The business dataset includes the first business data of the first business that has not been completed between the first subject and the second subject, the business quantity of at least one historical business between the first subject and the second subject, and at least one set of historical business data of the at least one historical business. The second subject is any subject that has a business relationship with the first subject, and the first business is any business that has not been completed between the first subject and the second subject. The prediction module is used to predict the duration of the first service based on the number of services and at least one item from the at least one set of historical service data, as well as the service start time in the first service data. The prediction module is also used to predict the end time of the first service based on the service start time and the service duration. The prediction module is also used to predict business risks for the first entity based on the end time of all unfinished business between the first entity and all entities with which it has business relationships.

11. An electronic device, characterized in that, include: processor; as well as Memory for storing the executable instructions of the processor; The processor is configured to execute the data processing method of any one of claims 1 to 9 by executing the executable instructions.

12. A non-transitory computer-readable storage medium, wherein when instructions in the storage medium are executed by a processor of a mobile terminal, the mobile terminal is enabled to perform the data processing method of any one of claims 1 to 9.

13. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the data processing method as described in any one of claims 1 to 9.

Images