Threshold generation method and device for anti-money laundering rule model

Abstract

The invention discloses an anti-money laundering rule model threshold generation method and device, and the method comprises the steps: firstly carrying out the feature extraction of transaction information, obtaining a plurality of feature indexes, carrying out the calculation of the plurality of feature indexes, and obtaining the index value of each feature index; sorting the index value of each feature index to obtain a corresponding index value set; setting an initial index quantile of the anti-money laundering rule model; according to the initial index quantile and the index value set, determining a serial number corresponding to a threshold value of the anti-money laundering rule model; and finally, according to the serial number corresponding to the threshold value, determining a corresponding target index value from the index value set, and taking the target index value as the threshold value of the anti-money laundering rule model. According to the technical scheme, the anti-money laundering rule model threshold value is dynamically generated through the statistical quantile knowledge, manual calculation is not needed, and the generation efficiency of the anti-money laundering rule model threshold value can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular to a threshold generation method and device for an anti-money laundering rule model. Background Art

[0002] With the continuous development of the financial market, anti-money laundering work has become increasingly important. At present, the main way for banks to identify suspicious transactions for anti-money laundering is to filter and identify massive transactions through rule models. The transaction flow is filtered through the rule model, and finally a part of the information that meets the characteristics of money laundering is screened out, forming an early warning for further processing. The rule model can be adjusted, updated and iteratively optimized according to business needs.

[0003] The performance of the anti-money laundering rule model depends largely on the threshold setting. The determination of the threshold in the model is usually based on expert experience to manually verify the model effects of different thresholds in different transaction scenarios, which consumes a lot of manpower and is ultimately confirmed through repeated verification. The process is time-consuming, resulting in low efficiency in generating the threshold of the anti-money laundering rule model. Summary of the invention

[0004] In view of this, the present application provides a method and device for generating a threshold value of an anti-money laundering rule model, aiming to improve the generation efficiency of the threshold value of the anti-money laundering rule model.

[0005] In a first aspect, the present application provides a threshold generation method for an anti-money laundering rule model, the method comprising:

[0006] Extracting features from the transaction information to obtain a plurality of feature indicators, and calculating each of the plurality of feature indicators to obtain an indicator value of each feature indicator;

[0007] Sorting the indicator value of each characteristic indicator to obtain a corresponding indicator value set, wherein the indicator value set includes a corresponding relationship between the indicator value and the sequence number;

[0008] Setting an initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set;

[0009] Determining a sequence number corresponding to a threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set;

[0010] According to the sequence number corresponding to the threshold value, a corresponding target indicator value is determined from the indicator value set, and the target indicator value is used as the threshold value of the anti-money laundering rule model.

[0011] Optionally, determining the sequence number corresponding to the threshold of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set includes:

[0012] Determining a total number of indicator values ​​according to the indicator value set;

[0013] According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

[0014] Optionally, setting a target warning quantity and a target warning confirmation rate of the anti-money laundering rule model;

[0015] After taking the target indicator value as the threshold of the anti-money laundering rule model, running the anti-money laundering rule model using the threshold to obtain an early warning result of the anti-money laundering rule model;

[0016] According to the warning result, the current warning quantity and the current warning confirmation rate of the anti-money laundering rule model are calculated;

[0017] The initial indicator quantile is adjusted according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

[0018] Optionally, adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes:

[0019] If the current warning amount is greater than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is lowered according to a preset step size.

[0020] Optionally, adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes:

[0021] If the current warning amount is less than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

[0022] Optionally, adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes:

[0023] If the current warning amount is less than the target warning amount, and the current warning confirmation rate is greater than or equal to the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

[0024] In a second aspect, the present application provides a threshold generation device for an anti-money laundering rule model, the device comprising:

[0025] A feature extraction module is used to extract features from transaction information to obtain multiple feature indicators, and to calculate each of the multiple feature indicators to obtain an indicator value of each feature indicator;

[0026] A sorting module, used to sort the index value of each characteristic index to obtain a corresponding index value set, wherein the index value set includes a corresponding relationship between the index value and the sequence number;

[0027] A setting module, used to set the initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set;

[0028] A determination module, used to determine the sequence number corresponding to the threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set;

[0029] The determination module is further used to determine a corresponding target indicator value from the indicator value set according to a sequence number corresponding to the threshold value, and use the target indicator value as a threshold value of the anti-money laundering rule model.

[0030] Optionally, the determining module is specifically configured to

[0031] Determining a total number of indicator values ​​according to the indicator value set;

[0032] According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

[0033] Optionally, the setting module is further used to set the target warning quantity and target warning confirmation rate of the anti-money laundering rule model;

[0034] After taking the target indicator value as the threshold of the anti-money laundering rule model, the device further includes:

[0035] An operation module, used to operate the anti-money laundering rule model using the threshold value to obtain an early warning result of the anti-money laundering rule model;

[0036] A calculation module, used for calculating the current warning quantity and current warning confirmation rate of the anti-money laundering rule model according to the warning result;

[0037] An adjustment module is used to adjust the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

[0038] Optionally, the adjustment module is specifically configured to lower the initial indicator quantile according to a preset step size if the current warning amount is greater than the target warning amount and the current warning confirmation rate is less than the target warning confirmation rate.

[0039] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising:

[0040] A memory for storing one or more programs;

[0041] Processor; when the one or more programs are executed by the processor, the threshold generation method of the anti-money laundering rule model described in any one of the first aspects is implemented.

[0042] In a fourth aspect, an embodiment of the present application provides a computer storage medium, in which a program is stored. When the program is executed by a processor, the threshold generation method of the anti-money laundering rule model described in any one of the first aspects is implemented.

[0043] The above technical solution has the following beneficial effects:

[0044] The present application provides a threshold generation method and device for an anti-money laundering rule model. When executing the method, firstly, feature extraction is performed on transaction information to obtain multiple feature indicators, and each feature indicator in the multiple feature indicators is calculated to obtain the indicator value of each feature indicator; then, the indicator value of each feature indicator is sorted to obtain a corresponding indicator value set, and the indicator value set includes the corresponding relationship between the indicator value and the sequence number; then, the initial indicator quantile of the anti-money laundering rule model is set, and the initial indicator quantile is used to determine the position of the threshold value in the indicator value set; then, according to the initial indicator quantile and the indicator value set, the sequence number corresponding to the threshold value of the anti-money laundering rule model is determined; finally, according to the sequence number corresponding to the threshold value, the corresponding target indicator value is determined from the indicator value set, and the target indicator value is used as the threshold value of the anti-money laundering rule model. Through the above technical scheme, the present application configures the indicator quantile for a large number of feature indicators through the concept of quantile, and dynamically generates the threshold value of the anti-money laundering rule model by using the knowledge of statistical quantiles, which does not require manual calculation, and can improve the generation efficiency of the threshold value of the anti-money laundering rule model. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 An example diagram of the model composition relationship provided in the embodiment of the present application;

[0047] Figure 2 A flow chart of a threshold generation method for an anti-money laundering rule model provided in an embodiment of the present application;

[0048] Figure 3 A schematic diagram of the structure of a threshold value generating device for an anti-money laundering rule model provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0050] In order to facilitate further understanding of the technical solution provided by this application, the background technology involved in this application will be explained below.

[0051] The composition of the anti-money laundering rule model is generally in the form of indicators, rules and models, where the indicator is the calculation result set of a certain data feature, each indicator is combined with a threshold to form a rule, and several rules are combined into a model, such as Figure 1 As shown, Figure 1 An example diagram of the model composition relationship provided in the embodiment of the present application.

[0052] The anti-money laundering rule model is generally designed and formulated with reference to relevant legal documents. The threshold in the model is usually determined on the basis of defining the indicators and calculating the indicator result set. The threshold is determined according to the data results through the analysis of the indicator result set. At the same time, the model is analyzed according to the final warning results, and the threshold of the relevant rules in the model is continuously adjusted. The final threshold information is obtained through repeated verification. In this method, parameter adjustment is a relatively time-consuming task in the model development process, which consumes a lot of manpower and is ultimately confirmed through repeated verification. The process is relatively time-consuming, resulting in low efficiency in generating the threshold of the anti-money laundering rule model.

[0053] In order to overcome the above technical problems, the embodiment of the present application provides a threshold generation method and device for an anti-money laundering rule model. When executing the method, firstly, feature extraction is performed on the transaction information to obtain multiple feature indicators, and each feature indicator in the multiple feature indicators is calculated to obtain the index value of each feature indicator; then the index value of each feature indicator is sorted to obtain a corresponding index value set, and the index value set includes the corresponding relationship between the index value and the sequence number; then the initial index quantile of the anti-money laundering rule model is set, and the initial index quantile is used to determine the position of the threshold value in the index value set; then, according to the initial index quantile and the index value set, the sequence number corresponding to the threshold value of the anti-money laundering rule model is determined; finally, according to the sequence number corresponding to the threshold value, the corresponding target index value is determined from the index value set, and the target index value is used as the threshold value of the anti-money laundering rule model. Through the above technical solution, the present application configures the index quantile for a large number of feature indicators through the concept of quantile, and dynamically generates the threshold value of the anti-money laundering rule model by using the knowledge of statistical quantiles, without manual calculation, which can improve the generation efficiency of the threshold value of the anti-money laundering rule model.

[0054] The technical terms of the embodiments of the present application are introduced as follows:

[0055] Quantile: Quantile is a concept in statistics that is used to describe the distribution of data. It refers to the value at a certain percentile after the data is sorted by size. Quantile is usually used to describe the distribution of data, abnormal values ​​and outliers.

[0056] Indicator: An indicator is usually used to describe the identity or value of a specific object.

[0057] Rule: A rule is a set of statements or conditions that indicate or describe a specific situation. It is used to determine whether a specific object has a certain characteristic. It consists of indicators, operators, and thresholds.

[0058] Rule model: A model is generated by one or more rules and the logical relationships between rules.

[0059] Threshold refers to the decision boundary used to distinguish different categories in the rule model. In the rule model, the threshold is usually used to convert continuous values ​​into discrete categories. The choice of the rule model threshold has a great impact on the performance of the model. If the threshold is set too low, the model may be too sensitive and many samples may be misclassified into a certain category; if the threshold is set too high, the model may be too conservative and many samples may be misclassified into another category. Therefore, choosing an appropriate rule model threshold is crucial to improving the accuracy and generalization ability of the model.

[0060] Warning quantity: the magnitude of the results generated by the model.

[0061] Warning confirmation rate: After the model warning results are manually processed, the proportion of confirmed results in the entire model warning results.

[0062] The method of the embodiment of the present invention may be executed by a threshold generation device of an anti-money laundering rule model, which may be implemented by software and / or hardware and may generally be integrated into a server or terminal device.

[0063] See also Figure 2 , Figure 2 A flowchart of a threshold generation method for an anti-money laundering rule model provided in an embodiment of the present application.

[0064] The method may include:

[0065] Step S201: extracting features from transaction information to obtain a plurality of feature indicators, and calculating each of the plurality of feature indicators to obtain an indicator value of each feature indicator.

[0066] In this step, the threshold value generating device of the anti-money laundering rule model first extracts features from the transaction information to obtain a plurality of feature indicators, and calculates each of the plurality of feature indicators to obtain an indicator value of each feature indicator.

[0067] Specifically, transaction information may include customer information and account information.

[0068] Feature indicators are obtained by extracting features from transaction information and represent different aspects or attributes of transaction information. For example, feature indicators may include transaction frequency, average transaction amount, diversity of transaction locations, etc.

[0069] Step S202: sorting the indicator value of each characteristic indicator to obtain a corresponding indicator value set, wherein the indicator value set includes a corresponding relationship between the indicator value and the sequence number.

[0070] Specifically, after obtaining the index value of each characteristic index, the index value of each characteristic index is sorted in reverse order from large to small to generate a set of corresponding relationships {INDIC_VAL, INDIC_NUM}, where INDIC_VAL refers to the generated index result value; INDIC_NUM represents the serial number of the index value in reverse order.

[0071] Step S203: setting an initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set.

[0072] It should be noted that the initial indicator quantile can be set based on experience, model characteristics, source data exploration, and the way the model is run in a trial environment, and the value range is any value in the interval [0,1].

[0073] Step S204: Determine the sequence number corresponding to the threshold of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set.

[0074] In this step, the threshold value generating device of the anti-money laundering rule model can determine the sequence number corresponding to the threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set.

[0075] Specifically, in a possible implementation, step S204 may include:

[0076] Determining a total number of indicator values ​​according to the indicator value set;

[0077] According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

[0078] In this embodiment, the calculation formula for determining the sequence number corresponding to the threshold of the anti-money laundering rule model may include:

[0079] INDIC_ROW=INDIC_SUM*(1-INDIC_LINE)

[0080] (THRESHOLD,INDIC_ROW)={INDIC_VAL,INDIC_NUM}

[0081] INDIC_ROW=INDIC_NUM

[0082] Where INDIC_ROW: the serial number corresponding to the threshold; INDIC_SUM: the total number of indicator values; INDIC_LINE: the indicator quantile; THRESHOLD: the threshold, the value of INDIC_VAL corresponding to INDIC_ROW=INDIC_NUM in the set described in step 1.

[0083] Step S205: According to the sequence number corresponding to the threshold, a corresponding target indicator value is determined from the indicator value set, and the target indicator value is used as the threshold of the anti-money laundering rule model.

[0084] In this step, after determining the sequence number corresponding to the threshold of the anti-money laundering rule model in step S204, the corresponding target indicator value is further determined from the indicator value set according to the sequence number corresponding to the threshold, and the target indicator value is used as the threshold of the anti-money laundering rule model.

[0085] It can be seen from the above technical scheme that the present application first extracts features from transaction information to obtain multiple feature indicators, calculates each feature indicator in the multiple feature indicators, and obtains the indicator value of each feature indicator; then sorts the indicator value of each feature indicator to obtain a corresponding indicator value set, and the indicator value set includes the corresponding relationship between the indicator value and the sequence number; then sets the initial indicator quantile of the anti-money laundering rule model, and the initial indicator quantile is used to determine the position of the threshold value in the indicator value set; then determines the sequence number corresponding to the threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set; finally, determines the corresponding target indicator value from the indicator value set according to the sequence number corresponding to the threshold value, and uses the target indicator value as the threshold value of the anti-money laundering rule model. Through the above technical scheme, the present application configures indicator quantiles for a large number of feature indicators through the concept of quantiles, and dynamically generates the threshold value of the anti-money laundering rule model using statistical quantile knowledge, which does not require manual calculation and can improve the generation efficiency of the threshold value of the anti-money laundering rule model.

[0086] Considering the existing technology, the threshold of the anti-money laundering rule model is adjusted by monitoring the quality of the model results, actively optimizing the model threshold or model logic to improve the model warning volume and warning confirmation rate. The warning level is directly related to the pressure of warning processing. Too many warnings will increase the pressure of warning processing personnel, and at the same time, many invalid warnings will be generated, and the warning confirmation rate will be reduced. If the number of warnings is too small, the model will miss warnings or the model will fail. At this stage, the modification of model thresholds needs to be optimized after expert monitoring and analysis of model results, which is not timely and inefficient.

[0087] Therefore, in a possible implementation, the method further includes: setting a target warning quantity and a target warning confirmation rate of the anti-money laundering rule model;

[0088] After taking the target indicator value as the threshold of the anti-money laundering rule model, running the anti-money laundering rule model using the threshold to obtain an early warning result of the anti-money laundering rule model;

[0089] According to the warning result, the current warning quantity and the current warning confirmation rate of the anti-money laundering rule model are calculated;

[0090] The initial indicator quantile is adjusted according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

[0091] Specifically, after the target indicator value is used as the threshold of the anti-money laundering rule model, the anti-money laundering rule model is run using the threshold to obtain the early warning result of the anti-money laundering rule model. According to the early warning result, the current early warning amount and the current early warning confirmation rate of the anti-money laundering rule model are calculated, and then the initial indicator quantile is adjusted according to the current early warning amount and the target early warning amount, and according to the current early warning confirmation rate and the target early warning confirmation rate.

[0092] It should be noted that the embodiment of the present application dynamically adjusts the quantiles to automatically correct the model thresholds based on the warning quantity and confirmation rate of the model, so that the model warning quantity and confirmation rate meet business expectations to the greatest extent possible.

[0093] It can be understood that the target warning amount and the target warning confirmation rate in the embodiment of the present application can be set according to actual needs and are not limited here.

[0094] In a possible implementation, adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes:

[0095] If the current warning amount is greater than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is lowered according to a preset step size.

[0096] If the current warning amount is less than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

[0097] If the current warning amount is less than the target warning amount, and the current warning confirmation rate is greater than or equal to the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

[0098] Specifically, the dynamic adjustment of the indicator quantiles in the model is achieved by comparing the current warning quantity and warning confirmation rate of the model with the set warning quantity and warning confirmation rate. Priority should be given to ensuring that the model warning quantity is greater than or equal to the set value. Otherwise, the warning quantity will be too small or even zero, resulting in missed warnings. The warning confirmation rate should be considered only when the warning quantity meets expectations.

[0099] If the current warning volume > the set warning volume and the current warning confirmation rate < the set warning confirmation rate, the quantile point is adjusted downward according to the step size;

[0100] If the current warning quantity is less than the set warning quantity and the current warning confirmation rate is less than the set warning confirmation rate, the quantile will be adjusted upward according to the step size;

[0101] If the current warning quantity is greater than or equal to the set warning quantity, and the current warning confirmation rate is greater than or equal to the current warning confirmation rate, the quantile remains unchanged; if the current warning quantity is less than the set warning quantity, and the current warning confirmation rate is greater than or equal to the set warning confirmation rate, the quantile is adjusted upward according to the step size.

[0102] It is understandable that the step length can be adjusted according to actual conditions and is not limited here.

[0103] It should be noted that the embodiment of the present application can quickly improve the model warning confirmation rate and improve the model accuracy through the daily dynamic generation of thresholds and the adaptive adjustment of quantile values.

[0104] For the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the order of the actions described, because according to the present invention, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0105] The above are some specific implementations of the threshold generation method of the anti-money laundering rule model provided in the embodiment of the present application. Based on this, the present application also provides a corresponding device. The threshold generation device of the anti-money laundering rule model provided in the embodiment of the present application will be introduced from the perspective of functional modularization.

[0106] See also Figure 3 , is a structural schematic diagram of a threshold value generating device of an anti-money laundering rule model provided in an embodiment of the present application, the device comprising:

[0107] The feature extraction module 301 is used to extract features from the transaction information to obtain multiple feature indicators, and calculate each of the multiple feature indicators to obtain an indicator value of each feature indicator;

[0108] A sorting module 302 is used to sort the index value of each characteristic index to obtain a corresponding index value set, wherein the index value set includes a corresponding relationship between the index value and the sequence number;

[0109] A setting module 303 is used to set an initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set;

[0110] A determination module 304, configured to determine a sequence number corresponding to a threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set;

[0111] The determination module 304 is further configured to determine a corresponding target indicator value from the indicator value set according to a sequence number corresponding to the threshold value, and use the target indicator value as a threshold value of the anti-money laundering rule model.

[0112] Optionally, the determining module is specifically configured to

[0113] Determining a total number of indicator values ​​according to the indicator value set;

[0114] According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

[0115] Optionally, the setting module is further used to set the target warning quantity and target warning confirmation rate of the anti-money laundering rule model;

[0116] After taking the target indicator value as the threshold of the anti-money laundering rule model, the device further includes:

[0117] An operation module, used to operate the anti-money laundering rule model using the threshold value to obtain an early warning result of the anti-money laundering rule model;

[0118] A calculation module, used for calculating the current warning quantity and current warning confirmation rate of the anti-money laundering rule model according to the warning result;

[0119] An adjustment module is used to adjust the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

[0120] Optionally, the adjustment module is specifically configured to lower the initial indicator quantile according to a preset step size if the current warning amount is greater than the target warning amount and the current warning confirmation rate is less than the target warning confirmation rate.

[0121] It can be seen from the above technical scheme that the embodiment of the present application first extracts features from transaction information to obtain multiple feature indicators, calculates each feature indicator in the multiple feature indicators, and obtains the indicator value of each feature indicator; then sorts the indicator value of each feature indicator to obtain a corresponding indicator value set, and the indicator value set includes the corresponding relationship between the indicator value and the sequence number; then sets the initial indicator quantile of the anti-money laundering rule model, and the initial indicator quantile is used to determine the position of the threshold value in the indicator value set; then, according to the initial indicator quantile and the indicator value set, determines the sequence number corresponding to the threshold value of the anti-money laundering rule model; finally, according to the sequence number corresponding to the threshold value, determines the corresponding target indicator value from the indicator value set, and uses the target indicator value as the threshold value of the anti-money laundering rule model. Through the above technical scheme, the present application configures indicator quantiles for massive feature indicators through the concept of quantiles, and dynamically generates the threshold value of the anti-money laundering rule model using statistical quantile knowledge, without the need for manual calculation, which can improve the generation efficiency of the threshold value of the anti-money laundering rule model.

[0122] An embodiment of the present application further provides an electronic device, comprising: a memory for storing one or more programs;

[0123] Processor; when the one or more programs are executed by the processor, the threshold generation method of the anti-money laundering rule model described in the above embodiment is implemented.

[0124] An embodiment of the present application further provides a computer storage medium, in which a program is stored. When the program is executed by a processor, the threshold generation method of the anti-money laundering rule model described in the above embodiment is implemented.

[0125] The "first" and "second" in the names such as "first" and "second" (if any) mentioned in the embodiments of the present application are only used as name identifiers and do not represent the first or second in order.

[0126] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0127] Those skilled in the art can understand that the flowchart shown in the figure is only an example in which the embodiments of the present application can be implemented, and the scope of application of the embodiments of the present application is not limited by any aspect of the flowchart.

[0128] In the several embodiments provided in the present application, it should be understood that the disclosed methods, devices and equipment can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of the device or unit can be electrical, mechanical or other forms.

[0129] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. In addition, the functional units in the various embodiments of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0130] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program codes.

[0131] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A threshold generation method for an anti-money laundering rule model, characterized in that: The method comprises: Extracting features from the transaction information to obtain a plurality of feature indicators, and calculating each of the plurality of feature indicators to obtain an indicator value of each feature indicator; Sorting the indicator value of each characteristic indicator to obtain a corresponding indicator value set, wherein the indicator value set includes a corresponding relationship between the indicator value and the sequence number; Setting an initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set; Determining a sequence number corresponding to a threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set; According to the sequence number corresponding to the threshold value, a corresponding target indicator value is determined from the indicator value set, and the target indicator value is used as the threshold value of the anti-money laundering rule model.

2. The method according to claim 1, characterized in that The step of determining the sequence number corresponding to the threshold of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set includes: Determining a total number of indicator values ​​according to the indicator value set; According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

3. The method according to claim 1, characterized in that The method further includes: setting a target warning quantity and a target warning confirmation rate of the anti-money laundering rule model; After taking the target indicator value as the threshold of the anti-money laundering rule model, running the anti-money laundering rule model using the threshold to obtain an early warning result of the anti-money laundering rule model; According to the warning result, the current warning quantity and the current warning confirmation rate of the anti-money laundering rule model are calculated; The initial indicator quantile is adjusted according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

4. The method according to claim 3, characterized in that: The adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes: If the current warning amount is greater than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is lowered according to a preset step size.

5. The method according to claim 3, characterized in that: The adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes: If the current warning amount is less than the target warning amount, and the current warning confirmation rate is less than the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

6. The method according to claim 3, characterized in that: The adjusting the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate, includes: If the current warning amount is less than the target warning amount, and the current warning confirmation rate is greater than or equal to the target warning confirmation rate, the initial indicator quantile is increased according to a preset step size.

7. A threshold generation device for an anti-money laundering rule model, characterized in that: The device comprises: A feature extraction module is used to extract features from transaction information to obtain multiple feature indicators, and to calculate each of the multiple feature indicators to obtain an indicator value of each feature indicator; A sorting module, used to sort the index value of each characteristic index to obtain a corresponding index value set, wherein the index value set includes a corresponding relationship between the index value and the sequence number; A setting module, used to set the initial indicator quantile of the anti-money laundering rule model, wherein the initial indicator quantile is used to determine the position of the threshold value in the indicator value set; A determination module, used to determine the sequence number corresponding to the threshold value of the anti-money laundering rule model according to the initial indicator quantile and the indicator value set; The determination module is further used to determine a corresponding target indicator value from the indicator value set according to a sequence number corresponding to the threshold value, and use the target indicator value as a threshold value of the anti-money laundering rule model.

8. The device according to claim 7, characterized in that The determination module is specifically used for Determining a total number of indicator values ​​according to the indicator value set; According to the initial indicator quantile and the total number of indicator values, a sequence number corresponding to the threshold of the anti-money laundering rule model is determined.

9. The device according to claim 7, characterized in that The setting module is also used to set the target warning quantity and target warning confirmation rate of the anti-money laundering rule model; After taking the target indicator value as the threshold of the anti-money laundering rule model, the device further includes: An operation module, used to operate the anti-money laundering rule model using the threshold value to obtain an early warning result of the anti-money laundering rule model; A calculation module, used for calculating the current warning quantity and current warning confirmation rate of the anti-money laundering rule model according to the warning result; An adjustment module is used to adjust the initial indicator quantile according to the current warning amount and the target warning amount, and according to the current warning confirmation rate and the target warning confirmation rate.

10. The device according to claim 8, characterized in that The adjustment module is specifically used to lower the initial indicator quantile according to a preset step size if the current warning amount is greater than the target warning amount and the current warning confirmation rate is less than the target warning confirmation rate.

Images