JP7828991B2 - Information processing device and information processing method - Google Patents

Description

This invention relates to a technology for determining the available credit limit for users in financial services.

As a technology for determining the credit limit available to users in financial services, for example, Patent Document 1 discloses a method for determining the risks associated with increasing the maximum allowable credit limit using a machine learning model.

Special Publication No. 2020-522832

The present invention aims to appropriately determine the available credit limit for users in financial services.

To solve the above problems, the present invention comprises: an acquisition unit that acquires user information relating to users of financial services; a first trained machine learning model with the user information as an explanatory variable and termination risk information relating to the risk of termination of the financial service by the provider of the financial service as the objective variable; a second trained machine learning model with the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as the objective variable; and a first determination unit that determines the usage limit that each user can use in the financial service using the user information acquired by the acquisition unit for each user, wherein the first determination unit determines for each user The present invention provides an information processing device in which, when the cancellation risk information obtained by inputting user information relating to the user into the first trained model means that the risk level is lower than the first level, and the usage amount information obtained by inputting user information relating to the user into the second trained model means that the usage limit for the user is increased, and the first level is determined using the percentage of users for whom the first decision unit has decided to increase the usage limit and whose financial service has been forcibly terminated, and the percentage of users for whom the first decision unit has not decided to increase the usage limit and whose financial service has not been forcibly terminated.
Furthermore, the present invention comprises: an acquisition unit that acquires user information relating to users of financial services; a first trained model obtained by machine learning with the user information as an explanatory variable and termination risk information relating to the risk of termination of the financial service by the provider of the financial service as the objective variable; a second trained model obtained by machine learning with the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as the objective variable; and a first determination unit that determines the usage limit that each user can use in the financial service using the user information acquired by the acquisition unit for each user, wherein the first determination unit determines the user information relating to the user in the first trained model The present invention provides an information processing device in which, when the cancellation risk information obtained by inputting the user information relating to the user means that the risk is lower than the first level, and the usage amount information obtained by inputting the user information relating to the user into the second trained model means that the usage limit of the user is increased, and the second level is determined using the percentage of the usage amount used by a user who has been determined by the first decision unit to increase their usage limit, in which case the usage amount used by a user who has been determined by the first decision unit to increase their usage limit, in which case the usage amount used, in which case the usage amount used by a user who has not been determined by the first decision unit to increase their usage limit, in which case the usage limit is increased.
Furthermore, the present invention comprises the steps of: a computer acquiring user information relating to users of financial services; a first trained model obtained by machine learning in which the computer uses the user information as an explanatory variable and termination risk information relating to the risk of the financial service being forcibly terminated as an objective variable; a second trained model obtained by machine learning in which the computer uses the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as an objective variable; and determining the usage limit that each user can use in the financial service using the user information acquired for each user, wherein in the determination step, the computer determines the usage limit that each user can use for the financial service The present invention provides an information processing method characterized in that, when the cancellation risk information obtained by inputting user information into the first trained model means that the risk level is lower than the first level, and the usage amount information obtained by inputting user information relating to the user into the second trained model means that the usage limit for the user is increased, and the first level is determined using the percentage of users for whom the financial service was forcibly terminated in the determination step for whom it was decided to increase the usage limit, and the percentage of users for whom the financial service was not forcibly terminated in the determination step for whom it was not decided to increase the usage limit.
Furthermore, the present invention comprises the steps of: a computer acquiring user information relating to users of financial services; a first trained model obtained by machine learning with the user information as an explanatory variable and termination risk information relating to the risk of the financial service being forcibly terminated as an objective variable; a second trained model obtained by machine learning with the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as an objective variable; and determining the usage limit that each user can use in the financial service using the user information acquired for each user, wherein in the determination step, the computer uses the user information relating to the user in the first trained model The provided information processing method is characterized in that, when the cancellation risk information obtained by inputting it into the first level of risk is lower than the first level of risk, and the usage amount information obtained by inputting user information about the user into the second trained model is higher than the amount at the second level, the usage limit of the user is increased, and the second level is determined using the percentage of the usage amount used by users who were determined to have their usage limit increased in the determination step , where the usage amount used did not exceed a certain percentage of the usage limit after the determination, and the percentage of the usage amount used by users who were not determined to have their usage limit increased in the determination step, where the usage amount exceeded a certain percentage of the usage limit.

According to this invention, users can appropriately determine the available credit limit for financial services.

This figure shows an example of the configuration of an information processing system 1 according to an embodiment of the present invention. This figure shows the hardware configuration of the information processing device 30 according to the same embodiment. This figure shows an example of the functional configuration of the information processing device 30 according to the present invention. This figure shows an example of data stored by the information processing device 30 according to the same embodiment. This figure shows an example of a graph illustrating the method for determining the threshold in the same embodiment. This is a table illustrating the error rate in the same embodiment. This figure shows an example of a graph illustrating the method for determining the threshold in the same embodiment. This is a table illustrating the error rate in the same embodiment. This is a flowchart showing the learning operation by the information processing device 30. This is a flowchart showing the operation of determining the usage limit by the information processing device 30. This figure shows an example of the functional configuration of the information processing device 30 according to a modified example. This figure shows an example of rules stored in the information processing device 30 related to the modified example. This table illustrates the boundaries of each risk group in the modified example. This figure shows an example of how the usage limits are changed according to each risk group in the modified example. This figure shows an example of a merchant's commission rate in a modified example.

[composition]
Figure 1 shows an example of an information processing system 1 according to an embodiment of the present invention. The information processing system 1 is a system for electronically settling payments in commercial transactions in which users purchase goods and services at physical stores or online stores. As shown in Figure 1, the information processing system 1 comprises a user terminal 10 used by users, a store terminal 20 used by stores, a payment system 40 which includes an information processing device 30 according to the present invention and performs processing related to electronic payments, and a communication network 2 which includes a wireless communication network or a wired communication network that connects these in a communicative manner.

The user terminal 10, the store terminal 20, and the information processing device 30 are all communication-capable computers. Specifically, the user terminal 10 is a portable computer, such as a smartphone, wearable device, or tablet. The store terminal 20 is a communication terminal operated by store staff or other personnel. In this embodiment, the payment system 40 is managed by a telecommunications carrier providing communication services to the user terminal 10. The electronic payment service provided by this telecommunications carrier to the user is a financial service known as a combined communication charge payment. However, the financial service in this invention is not limited to this; it is sufficient if there is an upper limit (i.e., a credit limit to the user) on the amount of money that can be used by the user in the financial service provided by the financial service provider. Note that Figure 1 shows one user terminal, one store terminal, and one information processing device, but there may be multiple instances of each.

Figure 2 shows the hardware configuration of the information processing device 30. Physically, the information processing device 30 is configured as a computer including a processor 3001, memory 3002, storage 3003, communication device 3004, input device 3005, output device 3006, and a bus connecting these devices. Each of these devices operates on power supplied from a battery (not shown). In the following description, the term "device" can be read as a circuit, device, unit, etc. The hardware configuration of the information processing device 30 may include one or more of the devices shown in Figure 2, or it may be configured without some of the devices. Furthermore, multiple devices with different enclosures may be connected via communication to constitute the information processing device 30.

Each function in the information processing device 30 is realized by loading predetermined software (programs) onto hardware such as the processor 3001 and memory 3002. The processor 3001 then performs calculations, controls communication by the communication device 3004, and controls at least one of the reading and writing of data in the memory 3002 and storage 3003.

The processor 3001, for example, controls the entire computer by running the operating system. The processor 3001 may consist of a central processing unit (CPU) including interfaces with peripheral devices, control units, arithmetic units, registers, etc. Furthermore, for example, a baseband signal processing unit or a call processing unit may be implemented by the processor 3001.

The processor 3001 reads programs (program code), software modules, data, etc., from at least one of the storage 3003 and the communication device 3004 into the memory 3002, and executes various processes accordingly. The program used is one that causes the computer to execute at least a part of the operations described later. The functional blocks of the information processing device 30 may be stored in the memory 3002 and implemented by control programs running on the processor 3001. Various processes may be executed by a single processor 3001, or they may be executed simultaneously or sequentially by two or more processors 3001. The processor 3001 may be implemented by one or more chips. The program may also be transmitted to the information processing device 30 via a telecommunications line.

The memory 3002 is a computer-readable recording medium and may consist of at least one of the following: ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), RAM (Random Access Memory), etc. The memory 3002 may also be called a register, cache, main memory, etc. The memory 3002 can store executable programs (program code), software modules, etc., for carrying out the method according to this embodiment.

The storage device 3003 is a computer-readable recording medium and may consist of at least one of the following: an optical disc such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (e.g., a compact disc, a digital multipurpose disc, a Blu-ray® disc), a smart card, flash memory (e.g., a card, stick, or key drive), a floppy® disk, or a magnetic strip. The storage device 3003 may also be called an auxiliary storage device.

The communication device 3004 is hardware (transceiver/receiver device) for communicating between computers via the communication network 2, and is also referred to as a network device, network controller, network card, communication module, etc.

The input device 3005 is an input device that accepts external input (e.g., a key, microphone, switch, button, camera, sensor, 2D code reader, etc.). The output device 3006 is an output device that outputs to the outside (e.g., a display, speaker, LED lamp, etc.). The input device 3005 and the output device 3006 may be configured as an integrated unit (e.g., a touch panel).

The processor 3001, memory 3002, and other devices are connected by a bus for information communication. The bus may be configured using a single bus, or different buses may be used for each device.

The information processing device 30 may include hardware such as a microprocessor, a digital signal processor (DSP), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), and an FPGA (Field Programmable Gate Array), and some or all of each functional block may be implemented by this hardware. For example, the processor 3001 may be implemented using at least one of these hardware components.

The user terminal 10 and the store terminal 20, like the information processing device 30 shown in Figure 2, are equipped with hardware for realizing a computer, including a processor, memory, storage, communication device, input device, output device, and a bus connecting them. However, the communication device in the user terminal 10 is hardware (transceiver/receiver device) for realizing wireless communication, such as communicating between computers via the communication network 2 according to a communication standard such as LTE (Long Time Evolution). This communication device may include high-frequency switches, duplexers, filters, frequency synthesizers, etc., to realize at least one of frequency division duplex (FDD) and time division duplex (TDD). Furthermore, the transmission and reception antennas, amplifier section, transmission/reception section, and transmission path interface may also be realized by this communication device. The transmission/reception section may be physically or logically separated into a transmission section and a reception section.

Figure 3 is a block diagram showing the functional configuration of the information processing device 30. In the information processing device 30, the processor 3001 reads programs and other data from the storage 3003 into the memory 3002 and executes them, thereby realizing the functions of the acquisition unit 31, the learning unit 32, the storage unit 33, and the first decision unit 34.

The acquisition unit 31 reads and acquires user information from the user database (DB) 50, which consists of user information about users of financial services. User information includes attributes such as the user's age, as well as information such as the details of the contract for the financial service, the user history of the financial service, and the usage limit. Note that the user attributes used here may include attributes other than age. The user DB 50 may be owned by the information processing device 30, the payment system 40, or a system or device other than the payment system 40.

The learning unit 32 performs machine learning using machine learning algorithms such as logistic regression or random forest, and generates a forced churn rate model corresponding to the first trained model in this invention, and a usage prediction model corresponding to the second trained model in this invention.

The forced churn rate model is generated using machine learning with training data where user information is the explanatory variable and churn risk information—information about the risk of forced churn of a financial service by the financial service provider (in this case, a telecommunications carrier)—is the dependent variable. Forced churn by the financial service provider refers to cases where the financial service is terminated at the provider's discretion, regardless of the user's intentions, for example, due to user fraud or breach of contract. Information regarding such forced churns is included in the user information mentioned above. In the forced churn rate model, explanatory variables include attributes such as each user's age, as well as their financial service usage history (e.g., usage amount, frequency of use, number of uses, location of use, etc.), and the dependent variable is whether or not each user's financial service was forcibly terminated. Note that user attributes used here may include attributes other than age.

The usage prediction model is generated using machine learning with training data where user information is the explanatory variable and usage amount information (the amount of money spent by users in financial services) is the dependent variable. In the usage prediction model, explanatory variables include attributes such as the user's age, as well as their financial service usage history (e.g., frequency of use, number of uses, location of use, etc.). The dependent variable is the amount of money spent by that user, specifically from the information included in their financial service usage history. Note that the user attributes used here may include attributes other than age.

The memory unit 33 stores the forced churn rate model and the usage prediction model generated by the learning unit 32.

The first decision unit 34 inputs the user's information to the forced cancellation rate model stored in the memory unit 33 to obtain cancellation risk information regarding the risk of forced cancellation of financial services for that user. Specifically, this cancellation risk information is obtained as a forced cancellation score, which represents the risk of forced cancellation of financial services. This forced cancellation score takes values between 0 and 1, for example, where 0 indicates the minimum possibility of forced cancellation (i.e., forced cancellation is unlikely), and 1 indicates the maximum possibility of forced cancellation (i.e., forced cancellation is inevitable).

Furthermore, the first decision unit 34 inputs the user's user information to the usage prediction model stored in the memory unit 33 to obtain usage amount information regarding the amount of money the user will spend on financial services. This usage amount information is obtained as a usage prediction score, which is a predicted value of the amount the user will spend on financial services. This usage prediction score takes values from 0 to 1, for example, where 0 means the usage amount is minimal (i.e., not spent at all), and 1 means the usage amount exceeds a certain percentage (e.g., 80% or more) of the usage limit.

Here, Figure 4 is a table showing the relationship between cancellation risk information (forced cancellation score), usage amount information (usage prediction score), and increases or decreases in the credit limit. This table is stored by the first decision unit 34. The first decision unit 34 increases the credit limit of a user if the cancellation risk information is lower than the first level of risk and the usage amount information is higher than the second level of amount. In the example in Figure 4, the credit limit of a user is increased when the cancellation risk information (forced cancellation score) obtained for a user is less than 0.2 (first level of risk) and the usage amount information (usage prediction score) is 0.3 (second level of amount) or higher, that is, when it is within the shaded area S. In this way, when increasing the usage limit for each user, the first decision unit 34 can consider increasing the user's usage limit by one level (for example, from 30,000 yen to 50,000 yen) within a range of multiple usage limit levels with different sizes (for example, three levels of 30,000 yen, 50,000 yen, and 80,000 yen). Note that in Figure 4, if the relationship between the cancellation risk information (forced cancellation score) and the usage amount information (usage prediction score) falls outside the shaded area S, the usage limit will not be changed.

The first level described above is determined using the percentage of users whose financial services were forcibly terminated after the first decision unit 34 decided to increase their credit limit, and the percentage of users whose financial services were not forcibly terminated after the first decision unit 34 decided not to increase their credit limit.

Figure 5 shows an example of a graph illustrating the method for determining the first level. In Figure 5, the horizontal axis represents the forced termination score, and the vertical axis represents a value called the error rate. As shown in Figure 6, even if a user is not subject to forced termination of financial services based on the forced termination score (i.e., has a forced termination score of 0 and is considered a normal user), there may be cases where the user's financial services are actually forcibly terminated (referred to as a forcibly terminated user). In this case, the user who actually became a forcibly terminated user may have had their credit limit increased because they were not predicted to be a forcibly terminated user, meaning the prior prediction was incorrect. Graph c1 in Figure 5 shows the relationship between the forced termination score and the percentage of cases where the credit limit was increased because the user was not predicted to be a forcibly terminated user (error rate). This error rate corresponds to the percentage of users whose financial services were forcibly terminated among those whose credit limit was determined to be increased by the first determination unit 34.

On the other hand, even if a user is predicted to have their financial services forcibly terminated based on the forced termination score (a "forced termination user"), there may be cases where the user's financial services were not forcibly terminated in reality (a "normal user"). In this case, the user's credit limit was reduced because they were predicted to be a forced termination user, even though they should not have had their credit limit reduced. This means the prior prediction was incorrect. Graph c2 in Figure 5 shows the relationship between the forced termination score and the percentage (error rate) of cases where a user's credit limit was reduced because they were predicted to be a forced termination user, even though they should not have had their credit limit reduced. This error rate corresponds to the percentage of users whose financial services were not forcibly terminated for whom the first decision unit 34 did not determine that their credit limit should be increased.

In this embodiment, the forced termination score at the intersection of graph c1 and graph c2 is set as the first level described above, taking into consideration the balance between the percentage of users whose financial services were forcibly terminated after the first decision unit 34 determined to increase their credit limit, and the percentage of users whose financial services were not forcibly terminated after the first decision unit 34 determined to increase their credit limit.

Next, the second level described above is determined using the percentage of users whose spending limit was increased by the first decision unit 34 and whose spending did not exceed a certain percentage of that increased spending limit, and the percentage of users whose spending was not increased by the first decision unit 34 and whose spending exceeded a certain percentage of that increased spending limit.

Figure 7 shows an example of a graph illustrating the second level determination method. In Figure 7, the horizontal axis represents the usage prediction score, and the vertical axis represents the error rate. As shown in Figure 8, even for users whose usage limit has been increased based on the usage prediction score, there may be cases where they did not actually use a certain percentage (e.g., 80% or more) of the determined usage limit. In such cases, increasing the usage limit was unnecessary, meaning the prior prediction was incorrect. Graph c1 in Figure 7 shows the relationship between the usage prediction score and the percentage of cases (error rate) where, despite an increase in the usage limit, a certain percentage (80% or more) of the usage limit was not used. This error rate corresponds to the percentage of users whose usage amount was not exceeded by a certain percentage of the determined usage limit by the first determination unit 34.

On the other hand, even for users whose spending limit was increased based on their spending prediction score, there may be cases where they actually used a certain percentage (e.g., 80% or more) of that limit. In such cases, an increase in the spending limit was necessary; in other words, the prior prediction was incorrect. Graph c2 in Figure 7 shows the relationship between the spending prediction score and the percentage (error rate) of cases where a certain percentage (80% or more) of the spending limit was used despite not having the limit increased. This error rate corresponds to the percentage of cases where the amount spent by users who were not determined to have their spending limit increased by the first decision unit 34 exceeded a certain percentage of that spending limit.

In this embodiment, the usage prediction score at the intersection of graph c3 and graph c4 is set to the second level described above, taking into account the balance between the percentage of users whose usage amount was not increased by the first determination unit 34 and exceeded a certain percentage of that usage limit, and the percentage of users whose usage amount was not increased by the first determination unit 34 and exceeded that certain percentage.

In Figure 4, it was explained that the limit is increased when the relationship between the cancellation risk information (forced cancellation score) and the usage amount information (usage prediction score) falls within the shaded area S. However, for example, the first decision unit 34 may reduce the usage limit for each user if the cancellation risk information is higher than the risk level of the third level and the usage amount information is lower than the amount of the fourth level. For example, in Figure 4, if the cancellation risk information (forced cancellation score) for a certain user is 0.5 (risk level 3) or higher and the usage amount information (usage prediction score) is less than 0.1 (amount of the fourth level), that is, when it falls within the shaded area V, the usage limit for that user may be reduced. When reducing the usage limit for each user in this way, the first decision unit 34 may reduce the user's usage limit by one level out of multiple usage limit levels of different sizes.

Furthermore, instead of the first decision unit 34 totaling or reducing the usage limit in stages as described above, it may, for example, reduce the usage limit for each user as the risk indicated by the cancellation risk information is greater, and increase the usage limit for each user as the amount indicated by the usage amount information is larger.

[Operation]
Next, the operation of the information processing device 30 will be described with reference to Figures 9 and 10. In the following description, when the information processing device 30 is described as the main unit of processing, it specifically means that processing is executed by loading predetermined software (programs) onto hardware such as the processor 3001 and memory 3002, causing the processor 3001 to perform calculations and controlling communication by the communication device 3004 and the reading and/or writing of data in the memory 3002 and storage 3003.

[Learning behavior]
First, the learning operation by the information processing device 30 will be explained. In Figure 9, the acquisition unit 31 acquires user information for multiple users from the user DB 50 (step S11).

Next, the learning unit 32 generates training data to be used in a predetermined machine learning algorithm using the user information acquired by the acquisition unit 31 (step S12). Then, the learning unit 32 generates a forced churn rate model and a usage prediction model by performing machine learning using the training data and the machine learning algorithm (step S13).

The memory unit 33 stores the forced churn rate model and the usage prediction model generated by the learning unit 32 (step S14).

[Determination of spending limit]
First, the operation of the information processing device 30 to determine the usage limit will be explained. The acquisition unit 31 acquires user information from the user DB 50 regarding the user for whom the usage limit is to be determined (step S21).

Next, the first decision unit 34 inputs the user information acquired by the acquisition unit 31, corresponding to the explanatory information for each model, into the forced cancellation rate model and the usage prediction model stored in the memory unit 33 (step S22).

Then, the first decision unit 34 determines the maximum amount of money that the user can use for financial services based on the information obtained from each model (step S23).

The usage limit determined by the first determination unit 34 is stored in the storage unit 33 as user information for that user (step S24).

According to the embodiments described above, by using the forced cancellation rate model and the usage prediction model, along with user information, it becomes possible to appropriately determine the available credit limit for financial services for each user.

[Variations]
The present invention is not limited to the embodiments described above. The embodiments described above may be modified as follows. Furthermore, two or more of the following modifications may be combined and implemented.
[Variation 1]
In the above embodiment, the user's usage limit was determined using a forced cancellation rate model (first trained model) and a usage prediction model (second trained model). In contrast, the information processing device 30 may be modified to include a second determination unit 35 and a setting unit 36, as shown in Figure 11. The second determination unit 35 does not use the forced cancellation rate model (first trained model) and the usage prediction model (second trained model), but rather determines the usage limit for each user using a rule that associates user information with the usage limit in a range where the usage limit is small (for example, 80,000 yen or less, referred to as the first range). Then, the first determination unit 24 determines the user's usage limit using the forced cancellation rate model (first trained model) and the usage prediction model (second trained model) in a range where the usage limit is larger than the first range (for example, more than 80,000 yen, referred to as the second range). In other words, for users with low spending limits, the spending limit is determined according to predetermined rules based on the user's attributes such as age and past payment history, while for users with high spending limits, the spending limit is determined using a machine learning algorithm.

When determining the usage limit using rules within the first scope, specifically, as shown in Figure 12, rules are used that associate each usage limit (e.g., 30,000 yen, 50,000 yen, 80,000 yen) with combinations of user information, such as the user's age, the contract period for the financial service, whether or not there have been any arrears in the past period, and the amount used in the past period. These rules are stored in the second determination unit 35.

[Modified example 2]
In the above modified example 1, the setting unit 36 may set the usage amount that forms the boundary between the first range and the second range based on the aforementioned forced cancellation rate score. Specifically, the setting unit 36 compares the forced cancellation rate score obtained for each user with the actual result of the financial service being canceled, and sets the boundary between the first range and the second range of the usage limit by setting a usage limit such that the degree of agreement between the two exceeds a certain threshold.

[Modification 3]
In the above modified example, users may be divided into multiple groups based on their forced cancellation score, and a different usage limit may be set for each group. Specifically, as shown in Figure 13, the setting unit 36 divides users into, for example, a low-risk score group with a relatively low forced cancellation score, a high-risk score group with a relatively high forced cancellation score, and a medium-risk score group whose forced cancellation score falls between the low-risk and high-risk score groups, according to their forced cancellation rate score. Then, as illustrated in the table in Figure 14, the first decision unit 34 does not change the usage limit for users belonging to the high-risk group, increases the usage limit for users belonging to the low-risk group, and determines a usage limit for users belonging to the medium-risk group according to their usage prediction score.

[Modification 4]
In the above modified example, the setting unit 36 may set the churn rate risk corresponding to the boundary between the high-risk group and the medium-risk group based on the average commission rate paid from the store where the user uses financial services to the provider of those services. In this case, the setting unit 36 stores the commission rate paid from each store (merchant) to the provider of financial services, as shown in Figure 15. For example, if the average commission rate paid to the provider of financial services is 4% of the amount used, the setting unit 36 sets the forced churn rate score corresponding to the boundary between the high-risk group and the medium-risk group to 0.04. Alternatively, the setting unit 36 may set the risk corresponding to the boundary between the medium-risk group and the low-risk group based on the minimum commission rate paid from the store where the user uses financial services to the provider of those services. For example, if the minimum commission rate paid to the provider of financial services is 1% of the amount used, the forced churn rate score corresponding to the boundary between the medium-risk group and the low-risk group is set to 0.01.

[Other variations]
The block diagrams used in the description of the above embodiments show functional units. These functional blocks (components) are realized by any combination of at least one of hardware and software. Furthermore, the method of realizing each functional block is not particularly limited. That is, each functional block may be realized using one device that is physically or logically coupled, or it may be realized using two or more physically or logically separated devices that are directly or indirectly connected (for example, using wired or wireless connections). A functional block may also be realized by combining software with the one or more of the above devices.

Functions include, but are not limited to, judgment, decision, determination, calculation, calculation, processing, derivation, investigation, exploration, confirmation, reception, transmission, output, access, resolution, selection, selection, establishment, comparison, assumption, expectation, assumption, broadcasting, notifying, communicating, forwarding, configuring, reconfiguring, allocating (mapping), and assigning. For example, a functional block (configuration part) that enables transmission is called a transmission unit or transmitter. As mentioned above, the implementation method is not particularly limited.

For example, an information processing device in one embodiment of this disclosure may function as a computer performing the processing described herein.

Each aspect/embodiment described in this disclosure may be applied to at least one of the following: LTE (Long Term Evolution), LTE-A (LTE-Advanced), SUPER 3G, IMT-Advanced, 4G (4th generation mobile communication system), 5G (5th generation mobile communication system), FRA (Future Radio Access), NR (new Radio), W-CDMA®, GSM®, CDMA2000, UMB (Ultra Mobile Broadband), IEEE 802.11 (Wi-Fi®), IEEE 802.16 (WiMAX®), IEEE 802.20, UWB (Ultra-WideBand), Bluetooth®, and other appropriate systems, as well as next-generation systems extended based thereon. Furthermore, multiple systems may be applied in combination (for example, a combination of at least one of LTE and LTE-A with 5G).

The present invention may also be an information processing method characterized by comprising the steps of: acquiring user information relating to users of financial services; generating a first trained model by machine learning with the user information as an explanatory variable and termination risk information relating to the risk of forced termination of the financial service as an objective variable; generating a second trained model by machine learning with the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as an objective variable; and determining the usage limit available to each user in the financial service using the user information acquired for each user. However, the processing procedures, sequences, flowcharts, etc., of each aspect/embodiment described in this disclosure may be rearranged in order, as long as there is no contradiction. For example, the method described in this disclosure presents elements of various steps using an exemplary order and is not limited to the specific order presented.

Input and output information may be stored in a specific location (e.g., memory) or managed using a management table. Input and output information may be overwritten, updated, or appended to. Output information may be deleted. Input information may be transmitted to other devices.

The determination may be made using a single-bit value (0 or 1), a Boolean value (true or false), or a numerical comparison (for example, a comparison with a predetermined value).

While the present disclosure has been described in detail above, it will be clear to those skilled in the art that the present disclosure is not limited to the embodiments described herein. The present disclosure can be implemented in modified and altered forms without departing from the intent and scope of the present disclosure as defined by the claims. Therefore, the descriptions in this disclosure are illustrative and not intended to be restrictive in any way.

Software, whether called software, firmware, middleware, microcode, hardware description language, or by any other name, should be broadly interpreted to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, execution threads, procedures, functions, etc. Furthermore, software, instructions, information, etc., may be transmitted or received via a transmission medium. For example, if software is transmitted from a website, server, or other remote source using at least one of wired technologies (such as coaxial cable, fiber optic cable, twisted pair, or digital subscriber line (DSL)) and wireless technologies (such as infrared or microwave), then at least one of these wired and wireless technologies is included within the definition of a transmission medium.

The information, signals, etc. described in this disclosure may be represented using any of the various different techniques. For example, the data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltage, current, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.
In addition, terms used in this disclosure and terms necessary for understanding this disclosure may be replaced with terms having the same or similar meaning.

Furthermore, the information, parameters, etc., described in this disclosure may be expressed using absolute values, relative values from a predetermined value, or corresponding other information.

In this disclosure, the phrase "based on" does not mean "based solely on" unless otherwise specified. In other words, the phrase "based on" means both "based solely on" and "based at least on."

Any reference to elements using designations such as “First,” “Second,” etc., as used in this disclosure, does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient way to distinguish between two or more elements. Therefore, references to the First and Second elements do not imply that only two elements may be employed, or that the First element must precede the Second element in any way.

In the above descriptions of the device configuration, "part" may be replaced with "means," "circuit," "device," etc.

Where the terms “include,” “including,” and their variations are used in this disclosure, these terms are intended to be inclusive, as is the term “comprising.” Furthermore, the term “or” as used in this disclosure is not intended to mean exclusive OR.

In this disclosure, if articles are added through translation, such as a, an, and the in English, this disclosure may include the fact that the noun following these articles is plural.

In this disclosure, the term "A and B are different" may mean "A and B are different from each other." Furthermore, the term may also mean "A and B are each different from C." Terms such as "separate" and "combine" may be interpreted similarly to "different."

1: Information processing system, 2: Communication network, 10: User terminal, 20: Store terminal, 30: Information processing device, 31: Acquisition unit, 32: Learning unit, 33: Storage unit, 34: First decision unit, 35: Second decision unit, 36: Setting unit, 3001: Processor, 3002: Memory, 3003: Storage, 3004: Communication device, 3005: Input device, 3006: Output device, 40: Payment system, 50: User database.

Claims (9)

An acquisition unit that acquires user information about users of financial services,
The system comprises: a first trained machine learning model with user information as an explanatory variable and cancellation risk information relating to the risk of cancellation of the financial service by the provider of the financial service as the dependent variable; a second trained machine learning model with user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as the dependent variable; and a first determination unit that determines the usage limit that each user can use in the financial service using the user information acquired by the acquisition unit for each user .
The first decision unit increases the user's usage limit if, for each user, the cancellation risk information obtained by inputting the user information for that user into the first trained model means that the cancellation risk information is lower than the first level of risk, and the usage amount information obtained by inputting the user information for that user into the second trained model means that the usage amount information is higher than the second level of amount.
The first level is determined using the percentage of users whose financial services were forcibly terminated by the first decision unit for whom the credit limit was increased, and the percentage of users whose financial services were not forcibly terminated by the first decision unit for whom the credit limit was not increased.
An information processing device characterized by the following: An acquisition unit that acquires user information about users of financial services,
A first trained machine learning model uses the user information as an explanatory variable and cancellation risk information relating to the risk of cancellation of the financial service by the provider of the financial service as the dependent variable; a second trained machine learning model uses the user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as the dependent variable; and a first determination unit uses the user information acquired by the acquisition unit for each user to determine the usage limit that the user can use in the financial service.
Equipped with,
The first decision unit increases the user's usage limit if, for each user, the cancellation risk information obtained by inputting the user information for that user into the first trained model means that the cancellation risk information is lower than the first level of risk, and the usage amount information obtained by inputting the user information for that user into the second trained model means that the usage amount information is higher than the second level of amount.
The information processing device is characterized in that the second level is determined using the percentage of the amount used by users whose usage limit has been increased by the first determination unit, in which the amount used did not exceed a certain percentage of the usage limit after the determination, and the percentage of the amount used by users whose usage limit has not been increased by the first determination unit, in which the amount used exceeded a certain percentage of the usage limit . The information processing device according to claim 1 or 2, wherein the first decision unit reduces the usage limit for each user when the cancellation risk information obtained by the first trained model means that the risk level is higher than the third level, and the usage amount information obtained by the second trained model means that the usage amount is lower than the amount at the fourth level. The information processing apparatus according to claim 1 or 2, characterized in that when the first determination unit increases the usage limit for each user, it increases the usage limit for that user by one level among multiple usage limit levels of different sizes. The information processing apparatus according to claim 1 or 2, characterized in that when the first determination unit reduces the usage limit for each user, it reduces the usage limit for that user by one level out of multiple usage limit levels of different sizes. The information processing apparatus according to claim 1 or 2, characterized in that the first decision unit reduces the usage limit of each user as the risk indicated by the cancellation risk information increases, and increases the usage limit of each user as the amount indicated by the usage amount information increases. The system includes a second determination unit that determines the credit limit for each user within a first range of the credit limit available to the user in financial services, using rules that associate information about the user with the credit limit.
The information processing apparatus according to claim 1 or 2, characterized in that the first determination unit determines the usage limit for each user in a second range in which the usage limit is greater than the first range, using the first trained model, the second trained model, and the user information. The computer obtains user information about users of financial services,
The computer comprises a first pre-trained machine learning model with user information as an explanatory variable and termination risk information relating to the risk of the financial service being forcibly terminated as an objective variable, a second pre-trained machine learning model with user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as an objective variable, and a step of determining the usage limit that the user can use in the financial service using the user information obtained for each user,
In the determination step described above, if the computer determines that for each user, the cancellation risk information obtained by inputting user information about that user into the first trained model is lower than the first level of risk, and the usage amount information obtained by inputting user information about that user into the second trained model is higher than the second level of amount, then the computer increases the usage limit for that user.
An information processing method characterized in that the first level is determined using the percentage of users whose financial service was forcibly terminated in the determination step for whom it was decided to increase the credit limit, and the percentage of users whose financial service was not forcibly terminated in the determination step for whom it was not decided to increase the credit limit. The computer obtains user information about users of financial services,
The computer comprises a first pre-trained machine learning model with user information as an explanatory variable and termination risk information relating to the risk of the financial service being forcibly terminated as an objective variable, a second pre-trained machine learning model with user information as an explanatory variable and usage amount information relating to the amount used by the user in the financial service as an objective variable, and a step of determining the usage limit that the user can use in the financial service using the user information obtained for each user,
In the determination step described above, if the computer determines that for each user, the cancellation risk information obtained by inputting user information about that user into the first trained model is lower than the first level of risk, and the usage amount information obtained by inputting user information about that user into the second trained model is higher than the second level of amount, then the computer increases the usage limit for that user.
The information processing method is characterized in that the second level is determined using the percentage of the amount used by users who have been determined to have their usage limit increased in the determination step , in which the amount used did not exceed a certain percentage of the usage limit after the determination, and the percentage of the amount used by users who have not been determined to have their usage limit increased in the determination step, in which the amount used exceeded a certain percentage of the usage limit.