JP6280368B2 - Information processing apparatus, information processing method, computer program, and recording medium - Google Patents

Description

  The present invention relates to a technique for determining a credit amount.

  In order to reduce the risk that credit card companies (hereinafter referred to as “card companies”) are unable to collect credit card (“card”) usage from customers, generally as one of the credits on the card Set usage limit (hereinafter referred to as “limit”). On the other hand, since the card company earns money by having the customer use the card, if the limit is set too low, the profit will also decrease. Patent Document 1 describes that revenues obtained from each customer are calculated from the amount used, the amount due, the amount collected, and the profit rate.

JP 2004-139198 A

  However, since the card usage modes vary from customer to customer, the limit calculated by the method described in Patent Document 1 in which the variety of card usage modes is not sufficiently considered is the highest in the profit of the card company. It is not necessarily the upper limit.

  Accordingly, an object of the present invention is to determine a credit amount that can improve profits of a card company or the like.

  An information processing apparatus according to an embodiment of the present invention acquires a usage amount probability distribution that associates a usage amount in a predetermined period with a probability that the usage amount is used in the predetermined period. Then, the information processing apparatus determines the usage limit amount based on the expected profit calculated by applying the usage limit amount that determines the limit of the usage amount in the predetermined period to the usage amount probability distribution.

  The information processing apparatus may determine the usage limit so that the expected profit is maximized.

  According to the present invention, it is possible to determine a credit amount that can improve profits of a card company or the like.

2 shows a configuration example of an information processing apparatus according to the present embodiment. The structural example of a usage history information table is shown. The structural example of a utilization characteristic information table is shown. An example of a usage amount probability distribution is shown. The structural example of a limit amount information table is shown. An example of calculating a limit amount that maximizes expected profit is shown. The flowchart example of limit amount determination processing is shown.

  Hereinafter, an example of processing for determining a credit card usage limit (credit amount) will be described with reference to the drawings.

  FIG. 1 shows a configuration example of hardware and functions of the information processing apparatus 10.

  The information processing apparatus 10 includes a CPU 11, a main memory 13, a storage unit 14, and the like, and these elements 11, 13, and 14 are connected by a bus 15 that can transmit data bidirectionally. The main memory 13 is configured by, for example, a DRAM (Dynamic Random Access Memory) or an MRAM (Magnetic Responsive Random Access Memory). The storage unit 14 includes, for example, an HDD (Hard Disk Drive) or an SSD (Solid State Drive).

  The information processing apparatus 10 realizes functions related to the usage characteristic information generation unit 21, the limit amount information generation unit 22, and the input / output unit 23 by executing a predetermined computer program on the CPU 11. The information processing apparatus 10 holds data related to the usage history information table 100, the usage characteristic information table 120, and the limit amount information table 140 in the storage unit 14.

  Next, the configuration of the usage history information table 100 will be described with reference to FIG.

  The usage history information table 100 holds usage history information as a record. The usage history information includes information related to the usage history of the customer's card. The usage history information includes a customer ID item 101, a monthly usage amount item 102, an attribute item 103, a contract period item 104, and an accident frequency item 105. The usage history information may include only some of these items or may include other items.

  The customer ID item 101 stores a customer ID, which is information for uniquely identifying a customer.

  The monthly usage amount item 102 stores the usage amount of the card for each past month of the customer corresponding to the customer ID. For example, the usage amount for each of January, February,..., December of 2012 is stored.

  The attribute item 103 stores information related to the customer attribute corresponding to the customer ID. The attributes include, for example, the customer's annual income, household composition, and age.

  The contract period item 104 stores the contract period of the customer's card corresponding to the customer ID.

  The number of accidents item 105 stores the number of accidents caused by the customer corresponding to the customer ID. An accident is, for example, a withdrawal failure or a late payment.

  The usage characteristic information generation unit 21 uses the usage history information to generate customer usage characteristic information and registers it in the usage characteristic information table 120. The usage characteristic information generation unit 21 includes a usage amount probability distribution generation unit 31 that generates a usage amount probability distribution.

  Next, the configuration of the usage characteristic information table 120 and the function of the usage characteristic information generation unit 21 will be described with reference to FIG.

  The usage characteristic information table 120 holds usage characteristic information as a record. The usage characteristic information includes information regarding the usage characteristics of the customer's card. The usage characteristic information includes a customer ID item 121, a rate item 122, a default rate item 123, a remaining bond rate item 124, a monthly usage forecast amount item 125, and a usage amount probability distribution item 126. Have The usage characteristic information may include only some of these items or may include other items.

  The customer ID item 121 is the same as the customer ID item 101 described in FIG.

  The rate item 122 stores a rate for the card usage of the customer corresponding to the customer ID. The rate may be a value corresponding to the card company's fee for the usage amount of the customer. In this embodiment, the annual rate is stored in the rate item 122, but the rate for another period may be stored. The rate may be the same regardless of the customer, or may be different for each customer. The usage characteristic information generation unit 21 may determine the rate of the customer based on the customer usage history information corresponding to the customer ID.

  The default rate item 123 stores a default rate indicating the probability of default occurring in the customer corresponding to the customer ID. Default means that the customer defaults. In the present embodiment, the default rate for the year is stored in the default rate item 123, but default rates for other periods may be stored. The usage characteristic information generation unit 21 may determine the default rate of the customer based on at least one of the attribute of the customer usage history information corresponding to the customer ID, the contract period, and the number of accidents. For example, the usage characteristic information generation unit 21 may increase the default rate as the customer has more accidents.

  The item 124 of the remaining bond rate stores a remaining bond rate indicating a ratio that cannot be collected from the customer when the customer corresponding to the customer ID has made a default. In other words, the remaining bond rate is the ratio of the amount that cannot be collected out of the amount lent to the customer when the customer defaults. The remaining bond rate may be the same regardless of the customer, or may be different for each customer. The usage characteristic information generation unit 21 may determine the remaining bond rate of the customer based on at least one of the attribute of the customer usage history information corresponding to the customer ID, the contract period, and the number of accidents. For example, the usage characteristic information generation unit 21 may lower the residual bond rate as the customer has a higher annual income in the attribute.

  The predicted monthly usage amount item 125 stores a predicted value of the card usage amount for one month of the customer corresponding to the customer ID. The usage characteristic information generation unit 21 may calculate a monthly usage predicted amount based on customer usage history information corresponding to the customer ID.

  For example, the usage characteristic information generation unit 21 may calculate the monthly usage predicted amount based on the monthly usage amount of the customer usage history information corresponding to the customer ID. The usage characteristic information generation unit 21 may calculate an average of monthly usage amounts of the usage history information corresponding to the customer ID for the last year for the monthly usage predicted amount.

  For example, the usage characteristic information generation unit 21 may calculate the monthly usage estimated amount based on the attribute of the usage history information corresponding to the customer ID. When the monthly usage amount does not exist or is insufficient as in the case of a new customer, the usage characteristic information generation unit 21 calculates a predetermined statistic for the monthly usage amount of another customer group having an attribute similar to that new customer. The monthly usage forecast amount of the new customer may be calculated by applying a conventional method.

  The usage amount probability distribution item 126 stores information on the usage amount probability distribution of the customer corresponding to the customer ID.

  The usage amount probability distribution is information indicating variation in the usage amount for each predetermined period (for example, every month) of the customer. That is, the usage amount probability distribution is information indicating which usage amount can be generated with what probability (frequency) in a predetermined period (that is, the usage amount occurrence probability). For example, the usage amount probability distribution generation unit 31 may generate a usage amount probability distribution of the customer by applying a predetermined statistical method to the monthly usage amount of the customer usage history information corresponding to the customer ID. .

  Next, the usage amount probability distribution will be further described with reference to the graph 200 of FIG. In the graph 200 of FIG. 4, a first usage amount probability distribution 201, a second usage amount probability distribution 202, and a third usage amount probability distribution 203 are plotted.

  The first usage probability distribution 201 has the highest probability that the monthly usage amount is 40,000 yen, and the probability decreases as the usage amount increases and decreases around the usage amount. For example, a customer who uses a card to pay a monthly public bill and does not use a card much other than that tends to have a probability distribution like the first usage amount probability distribution 201. A usage amount probability distribution in which the probability of a specific usage amount tends to be relatively high is referred to as a fixed amount usage distribution model.

  Such a flat-rate usage distribution model may be expressed by a normal distribution. A flat-rate usage model expressed by a normal distribution is called a normal distribution model. The normal distribution model may be expressed as a normal distribution function having the average value μ and the standard deviation σ as parameters. The average value μ and the standard deviation σ may be calculated based on the customer's past monthly usage amount.

  For example, the usage amount probability distribution generation unit 31 has an average value μ = 40,000 yen (predicted monthly usage amount) and a standard deviation σ = 100,000 yen based on the past monthly usage amount of the customer with the customer ID “1”. A normal distribution model is generated and associated with the customer. That is, the usage amount probability distribution generation unit 31 stores “normal distribution, μ = 40,000 yen, σ = 100,000 yen” in the item 125 of the usage amount probability distribution in the usage characteristic information of the customer ID “1”. That is, this normal distribution means that the monthly usage amount falls within the range of 40,000 yen ± 100,000 yen with a probability of about 95%.

  In the second usage amount probability distribution 202, the probability decreases as the usage amount increases. For example, a customer who does not use a card for periodic payment but uses a card for daily shopping or occasional high-priced shopping has a probability distribution like the second usage amount probability distribution 202. easy. The usage amount probability distribution in which the probability tends to decrease as the usage amount increases is referred to as a general usage distribution model.

  In the third usage amount probability distribution 203, although the probability decreases as the usage amount increases, the probability of a specific 200,000 yen is relatively high. For example, a customer who uses a card for daily shopping or the like and uses a card for a high payment (for example, payment of insurance premium) every six months is a probability distribution like the third usage probability distribution 203. It is easy to become. Although the probability tends to decrease as the usage amount increases, the usage amount probability distribution with a relatively high probability of the specific usage amount is referred to as a specific usage distribution model.

  In the usage amount probability distribution item 125 of the usage characteristic information table 120, any of the above usage distribution models may be stored, or usage distribution models indicating other characteristics may be stored.

  Here, it is assumed that the limit amount is set to 150,000 yen in the graph 200 of FIG. 4 (reference numeral 210). In this case, the customer having the third usage amount probability distribution 203 cannot use the card for payment of the insurance premium of 200,000 yen every six months. This is because the payment amount of 200,000 yen exceeds the limit of 150,000 yen. That is, when the card company sets the limit amount to 150,000 yen for the customer having the third usage amount probability distribution 203, the card company increases to 200,000 yen every six months that should be obtained when the limit amount is not set We cannot get such income. In other words, the card company may be able to improve profits by setting an appropriate limit based on the customer usage probability distribution.

  The limit amount information generation unit 22 generates customer limit amount information using the usage characteristic information and registers it in the limit amount information table 140. The limit amount information generation unit 22 includes an expected return calculation unit 32 that calculates expected return that is expected to be obtained from the customer, and a limit amount determination unit 33 that determines the limit amount of the customer.

  Next, the configuration of the limit amount information table 140 and the function of the limit amount information generation unit 22 will be described with reference to FIG.

  The limit amount information table 140 holds limit amount information as a record. The limit amount information includes information on the limit amount of the customer's card usage. The limit amount information includes a customer ID item 141, a current limit item 142, a provisional limit item 143, an expected income item 144, an expected expenditure item 145, and an expected income item 146. Have The limit amount information may include only some of these items or may include other items.

  The item 141 of customer ID is the same as the item 101 of customer ID demonstrated in FIG.

  The current limit item 142 stores a limit amount currently set for the customer corresponding to the customer ID.

  The provisionally set limit item 143 stores a provisionally set limit amount used for calculating expected profit. In the provisional limit amount item 143, a plurality of limit amounts that can be set by the card company may be stored, or a limit amount that maximizes expected profit, which will be described later, may be stored.

  The expected income item 144 stores the amount of income that the card company is expected to obtain from the customer corresponding to the customer ID when the provisional limit is applied. In this embodiment, the expected revenue item 144 stores the expected revenue for the year, but may also store the expected revenue for other predetermined periods.

  The expected expenditure item 145 stores an amount of expenditure with a risk that the card company may suffer a loss from the customer corresponding to the customer ID when the provisional limit is applied. In this embodiment, the expected expenditure item 145 stores the annual expected expenditure, but other expected expenditures may be stored.

  The expected profit item 146 stores the amount of profit (or profit / loss) expected to be received from the customer corresponding to the customer ID by the card company when the provisional limit is set. For example, the expected revenue calculation unit 32 may calculate “expected revenue = expected revenue−expected expenditure”.

  Revenue that does not take into account the limit amount can be calculated by “annual usage forecast amount × fee rate”, but expected income that takes into account the limit amount cannot be calculated simply as such. This is because when the limit amount is set, the customer cannot use the card to pay for a product or the like whose amount exceeds the limit amount. That is, in the usage amount probability distribution, the probability that a usage amount exceeding the limit amount occurs is “0”. Therefore, if a card company sets a relatively low limit for a customer with a usage probability distribution that has a relatively high probability of occurrence of usage exceeding the limit, the usage exceeding the limit will be If it occurs, you lose the opportunity to earn. The lost use of this revenue opportunity is called the opportunity loss.

  The expected revenue calculation unit 32 calculates the monthly usage prediction amount (referred to as “monthly expected usage prediction amount”) in consideration of the limit amount as “monthly expected usage prediction amount = monthly usage prediction amount−month opportunity loss amount”. . Then, the expected revenue calculation unit 32 calculates the expected revenue for one year as “expected revenue = expected monthly usage expected amount × 12 months × charge rate”.

  For example, assuming that the usage amount probability distribution is a normal distribution model, assuming that the monthly usage predicted amount is x, the monthly usage standard deviation is σ, and the limit amount is CL, the expected revenue calculation unit 32 calculates the monthly Expected usage forecast amount Y is calculated.

  Here, Φ (t) is a standard normal distribution function shown below.

  There are various possibilities for how a customer who wants to use more than the limit will actually act. For example, a customer who pays the amount up to the upper limit of the monthly usage amount with a card, or a customer who pays with a card in the next month for an amount exceeding the limit of the monthly usage amount is assumed. The expected profit calculation unit 32 may calculate the expected profit in consideration of the behavioral characteristics of these customers. The above-mentioned formulas 1 to 4 assume a customer who pays a card up to the upper limit of the monthly usage amount.

  For example, it is assumed that the usage probability distribution is a usage distribution model according to f (x), and a customer who has an opportunity to pay a usage amount exceeding the limit CL with a card gives up all the payments (for example, settles all with another card). ), The expected revenue calculation unit 32 calculates the expected monthly usage amount Y according to the following formula 5.

  For example, when it is assumed that a customer who has the opportunity to pay an amount exceeding the limit CL with a card pays the amount up to the upper limit of the monthly usage amount with the card, the expected revenue calculation unit 32 uses the following formula 6 To calculate the expected monthly usage Y.

  The expected profit calculation unit 32 calculates “expected expenditure = limit amount × residual bond rate × default rate” as the expected expenditure for one year when the card cannot be used after occurrence of default. The remaining bond rate may be determined based on the percentage of the total limit that is used by the customer when default occurs. Note that the remaining bond rate may be determined based on a ratio that can be finally collected when a default occurs. In this case, the expected expenditure may include the cost required for the collection operation.

Then, the expected revenue calculation unit 32 calculates “expected revenue = expected revenue−expected expenditure”. Here, since the expected income and the expected expenditure are values that can be changed according to a set limit amount, the expected revenue is also a value that can be changed depending on the set limit amount. Therefore, the card company can maximize the profit expected to be obtained from the customer by setting the limit amount so that the expected profit is maximized.

  Next, a specific example of calculating a limit amount that maximizes the expected profit will be described with reference to a graph 300 in FIG.

  A customer usage probability distribution 301 shown in the graph 300 is a normal distribution model, and it is assumed that an average value (monthly usage prediction amount) μ = 40,000 yen and a monthly usage standard deviation σ = 100,000 yen. In addition, the customer's rate is 2% and the remaining bond rate is 80%.

As shown in FIG. 6 (A), when the provisional limit is set to 100,000 yen, the expected profit A is calculated as follows.
・ 40,000 yen (estimated monthly usage)-16,000 yen (opportunity loss) = 24,000 yen (expected monthly expected usage)
・ 24,000 yen (expected monthly expected usage) x 12 (months) x 2% (rate) = 5780 yen (expected revenue)
・ 100,000 yen (provisional limit) x 80% (residual bond rate) x 2% (default rate) = 1600 yen (expected expenditure)
・ 5780 yen (expected revenue)-1600 yen (expected expenditure) = 4180 yen (expected revenue A)

  Here, the opportunity loss amount of 16,000 yen corresponds to the area 303A in FIG. 6A, and the monthly expected usage predicted amount of 24,000 yen corresponds to the area 302A in FIG. 6A. That is, the expected monthly usage amount is a value obtained by integrating the usage amount probability distribution 301 with respect to the usage amount “−∞ to 100,000 yen (temporary setting limit amount)”, and the opportunity loss amount is the usage amount probability distribution 301. Is a value obtained by integrating the usage amount “100,000 yen (provisional limit amount) to ∞”. The same applies to FIGS. 6B and 6C below.

When the provisional limit is set to 200,000 yen as shown in FIG. 6B, the expected profit B is calculated as follows.
・ 40,000 yen (estimated monthly usage)-22,000 yen (opportunity loss) = 38,000 yen (expected monthly expected usage)
・ 38,000 yen (expected monthly expected usage) x 12 (months) x 2% (rate) = 9116 yen (expected revenue)
・ 200,000 yen (provisional limit) x 80% (residual bond rate) x 2% (default rate) = 3200 yen (expected expenditure)
・ 9116 yen (expected revenue)-3200 yen (expected expenditure) = 5916 yen (expected revenue B)

When the provisional limit is set to 300,000 yen as shown in FIG. 5C, the expected return C is calculated as follows.
・ 40,000 yen (estimated monthly usage)-1010,000 yen (loss of opportunity loss) = 39,000 yen (expected monthly expected usage)
・ 3.99 million yen (expected monthly expected usage) x 12 (months) x 2% (rate) = 9573 yen (expected revenue)
・ 300,000 yen (limit) x 80% (residual bond rate) x 2% (default rate) = 4800 yen (expected expenditure)
・ 9573 yen (expected revenue)-4800 yen (expected expenditure) = 4773 yen (expected revenue C)

  Therefore, since “expected profit B> expected profit C> expected profit A”, the expected profit is maximized when “200,000 yen” is set as the limit among the above three provisional limits. In other words, the card company is likely to obtain the maximum profit from this customer by setting the limit amount of this customer to 200,000 yen.

  Therefore, the limit determining unit 33 determines the provisional limit “200,000 yen” that maximizes the expected profit as the limit for the customer.

  Note that the limit determining unit 33 does not necessarily have to determine the provisional limit that maximizes the expected profit as the limit. For example, when there is a possibility that the usage amount may be larger than expected due to external factors such as the implementation of a usage promotion campaign, the limit amount determination unit 33 may sacrifice the profit amount somewhat (that is, take a large risk). And a relatively large temporary limit may be determined as the limit. The limit amount determination unit 33 may determine the limit amount from among the temporarily set limit amounts included within a predetermined range from the maximum profit amount. For example, the limit amount determination unit 33 may determine the limit amount from the range from the maximum revenue amount to the revenue amount that pushes down the maximum revenue amount by 1%.

  The limit amount determination unit 33 may update the current limit amount item 142 of the limit amount information corresponding to the customer in the limit amount information table 140 to the determined limit amount.

  The input / output unit 23 may display various information included in the usage history information table 100, the usage characteristic information table 120, and the limit amount information table 140 on the display device. The input / output unit 23 may display the usage amount probability distribution of the customer in a graph format as shown in FIG. The input / output unit 23 may accept input of various parameters such as a charge rate, a default rate, and a residual bond rate from a user of the information processing apparatus 10. The input / output unit 23 may accept the selection of a target customer whose limit is to be calculated. The input / output unit 23 may display a plurality of provisional limit amounts and accept selection of a limit amount to be adopted from the user of the information processing apparatus 10.

  When the customer's limit is increased, the input / output unit 23 may notify the customer that the limit has been increased by e-mail, letter, FAX, or the like. This is because the size of the limit is regarded as status evaluation, and the customer's willingness to use the card may be improved. In this case, the expected revenue calculation unit 32 may calculate the expected revenue by adding the effect of increasing the customer's willingness to use the card by increasing the limit.

  Next, an example of processing for calculating the limit amount will be described with reference to the flowchart of FIG.

  The user of the information processing apparatus 10 or a predetermined program selects a customer ID (referred to as a target customer ID) for which a limit amount is to be calculated (S101).

  The usage amount probability distribution generation unit 31 generates a usage amount probability distribution based on the usage history information corresponding to the target customer ID, and registers it in the usage characteristic information table 120 (S102). Instead of generating a usage amount probability distribution, the usage amount probability distribution generation unit 31 acquires a usage distribution model that matches the target customer ID from a plurality of existing usage distribution models, and registers it in the usage characteristic information table 120. Also good.

  Based on the usage history information corresponding to the target customer ID, the usage characteristic information generation unit 21 calculates usage rates, a default rate, a remaining bond rate, and a monthly usage prediction amount, generates usage characteristic information, and uses characteristic information table 120. (S103). The usage characteristic information generation unit 21 may adopt a predetermined value and register it in the usage characteristic information table 120 instead of calculating the rate, the default rate, and the remaining bond rate.

  The expected revenue calculation unit 32 extracts the usage characteristic information corresponding to the target customer ID from the usage characteristic information table 120, and calculates the provisional limit amount and the expected profit based on the usage characteristic information (S104).

  The limit amount determination unit 33 determines the temporary limit amount that maximizes the expected profit as the limit amount corresponding to the target customer ID (S105). The limit amount determination unit 33 may present a plurality of temporarily set limit amounts to the user and allow the user to determine the limit amount instead of determining the temporarily set limit amount with the maximum expected profit as the limit amount. . The limit amount information generation unit 22 generates limit amount information corresponding to the target customer ID and registers it in the limit amount information table 140.

  The input / output unit 23 displays limit amount information corresponding to the target customer ID on the display device, and accepts, for example, correction or change of the limit amount from the user (S106).

  With the above processing, it is possible to set a limit amount at which revenue can be maximized for each customer. In other words, by setting an appropriate limit for the customer in consideration of the variety of card usage patterns of the customer, the profit of the card company can be improved.

  The above-described embodiments of the present invention are examples for explaining the present invention, and are not intended to limit the scope of the present invention only to those embodiments. Those skilled in the art can implement the present invention in various other modes without departing from the gist of the present invention.

  For example, some or all of the tables 100, 120, and 140 may be stored in another device, and the information processing device 10 may access the tables 100, 120, and 140 through a predetermined communication network. Further, the information processing apparatus 10 may transmit / receive input / output information related to the input / output unit 23 to / from another apparatus through a predetermined communication network.

  For example, the information processing apparatus 10 may be operated by a company different from the card company. That is, the information processing apparatus 10 does not hold the usage history information table 100 itself, receives customer usage history information from the card company, calculates an appropriate limit for the customer, and transmits the calculation result to the card company. May be.

  In this embodiment, one usage distribution model is associated with a customer. However, a plurality of different usage distribution models may be associated with a customer, and each usage distribution model may be associated with a probability that can be the usage distribution model. . Then, the expected revenue calculation unit 32 may calculate the expected revenue based on Bayesian estimation from the probabilities associated with each of the plurality of usage distribution models. Similarly, the expected return may be calculated based on Bayesian estimation from a plurality of different default rates associated with the customer.

  The limit amount information generating unit 22 appropriately categorizes customers based on the monthly usage forecast amount, standard deviation, default rate, and the like instead of calculating the limit amount using the above-described calculation formula for each of all customers. A limit amount may be calculated in units and registered in a predetermined limit amount calculated table. In that case, the limit amount information generating unit 22 may acquire the limit amount corresponding to the customer from the limit amount calculated table.

  For example, the use probability distribution item 126 may store different use probability distributions for each month in the future year for the customer ID. In this case, the expected revenue calculation unit 32 may calculate different expected revenues for each month using different usage probability distributions for each month.

  For example, the usage amount probability distribution item 126 may store a plurality of different usage amount probability distributions in consideration of increase / decrease in sales due to changes in economic conditions. In this case, the expected profit calculation unit 32 may calculate the expected profit using the usage amount probability distribution corresponding to the economic situation at that time.

  In the present embodiment, the calculation of the limit amount of the card has been described, but it can also be applied to the calculation of the limit amount other than the card. In other words, the present embodiment can be applied to various business fields in which the expected profit of the business operator can fluctuate according to the setting of the limit amount for the customer.

DESCRIPTION OF SYMBOLS 10 ... Information processing apparatus 21 ... Usage characteristic information generation part 22 ... Limit amount information generation part 31 ... Usage probability distribution generation part 32 ... Expected profit calculation part 33 ... Limit amount determination part 100 ... Usage history information table 120 ... Usage characteristic information table 140 ... Limit information table

Claims (12)

An information processing apparatus for determining a credit limit,
A storage unit that holds a usage amount probability distribution that associates a usage amount in a predetermined period with a probability that the usage amount is used in the predetermined period;
A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. And a limit determining unit to determine ,
The usage probability distribution is associated with each customer,
The limit amount determination unit determines the use limit amount for each of the customers using the use amount probability distribution associated with the customer . The information processing apparatus according to claim 1, wherein the limit amount determining unit determines the use limit amount so that the expected profit is maximized or equal to or greater than a predetermined value. The expected revenue is a value according to a difference between an expected income calculated based on the usage amount probability distribution and the usage limit amount and an expected expenditure calculated based on the usage limit amount,
The expected income is a value that increases according to the usage amount probability distribution when the usage limit increases.
The information processing apparatus according to claim 2, wherein the expected expenditure is a value that increases according to the usage limit when the usage limit increases. The information processing apparatus according to claim 3, wherein the expected income is a value according to an integrated value from a predetermined usage amount smaller than the usage limit amount to the usage limit amount according to the usage amount probability distribution. An information processing apparatus for determining a credit limit,
A storage unit that holds a usage amount probability distribution that associates a usage amount in a predetermined period with a probability that the usage amount is used in the predetermined period;
A usage limit that defines a limit on the usage amount for the predetermined period, and the expected return is a maximum based on an expected return that is a return expected when the usage limit is applied to the usage probability distribution. Or a limit determining unit that determines the above-mentioned limit of use so as to be equal to or greater than a predetermined value.
With
The expected revenue is a value according to a difference between an expected income calculated based on the usage amount probability distribution and the usage limit amount and an expected expenditure calculated based on the usage limit amount,
The expected income is a value that increases according to the usage amount probability distribution when the usage limit increases.
The expected expenditures, and the credit limit, the remaining bonds which is a ratio of recovery is disabled when the default, Ru value der according to the product of the default rate is the probability of default occurs, information processing apparatus . The information processing apparatus according to claim 1 , further comprising: a usage characteristic information generation unit configured to generate the usage amount probability distribution corresponding to the customer based on a usage amount history of the customer for each predetermined period. An information processing method for determining a credit limit,
A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. In an information processing method in which a computer executes determination ,
The usage probability distribution is associated with each customer,
The computer determines the usage limit for each customer using the usage probability distribution associated with the customer.   An information processing method for determining a credit limit,
  A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
  A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. decide
In an information processing method in which a computer executes
  The expected revenue is a value according to a difference between an expected income calculated based on the usage amount probability distribution and the usage limit amount and an expected expenditure calculated based on the usage limit amount,
  The expected income is a value that increases according to the usage amount probability distribution when the usage limit increases.
  The expected expenditure is a value according to the product of the usage limit amount, a residual bond rate that is a ratio that cannot be recovered when defaulted, and a default rate that is a probability of occurrence of default. A computer program for determining a credit limit,
A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. In a computer program for causing a computer to execute determination ,
The usage probability distribution is associated with each customer,
The usage limit amount for each customer is determined using the usage probability distribution associated with the customer.
To the computer.   A computer program for determining a credit limit,
  A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
  A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. decide
In a computer program for causing a computer to execute
  The expected revenue is a value according to a difference between an expected income calculated based on the usage amount probability distribution and the usage limit amount and an expected expenditure calculated based on the usage limit amount,
  The expected income is a value that increases according to the usage amount probability distribution when the usage limit increases.
  The expected expenditure is a value according to the product of the usage limit amount, a residual bond rate that is a ratio that cannot be recovered when defaulted, and a default rate that is a probability of occurrence of default. A recording medium on which a computer program for determining a credit limit is recorded,
A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. In a computer program for causing a computer to execute determination ,
The usage probability distribution is associated with each customer,
The usage limit amount for each customer is determined using the usage probability distribution associated with the customer.
Causing the computer to execute
A computer- readable recording medium having a computer program recorded thereon.   A recording medium on which a computer program for determining a credit limit is recorded,
  A usage amount probability distribution that associates the usage amount in a predetermined period with the probability that the usage amount is used in the predetermined period;
  A usage limit that defines a limit on the usage amount for the predetermined period, and is based on an expected return that is an expected return when the usage limit is applied to the usage probability distribution. decide
In a computer program for causing a computer to execute
  The expected revenue is a value according to a difference between an expected income calculated based on the usage amount probability distribution and the usage limit amount and an expected expenditure calculated based on the usage limit amount,
  The expected income is a value that increases according to the usage amount probability distribution when the usage limit increases.
  The expected spending is a value according to the product of the limit of use, a residual bond rate that is a ratio that cannot be recovered when defaulted, and a default rate that is a probability of occurrence of default.
A computer-readable recording medium having a computer program recorded thereon.

Images