Information processing device and information processing program

Abstract

This allows for the determination of coupon distribution patterns for multiple target groups, taking budget into consideration. [Solution] The information processing device of the embodiment comprises a calculation means, a selection means, a control means, and a first generation means. The calculation means calculates an effect amount that serves as an indicator of the effectiveness of each of a plurality of distribution patterns for distributing coupons to a plurality of customers within a budget. The selection means selects one of the plurality of distribution patterns based on the effect amount calculated by the calculation means. The control means controls the calculation means and the selection means to perform the calculation of expected profits and the selection of distribution patterns with respect to different budgets. The first generation means generates information that shows the plurality of distribution patterns selected by the selection means under the control of the control means in a manner that allows for comparison of the magnitude of the effect amounts calculated by the calculation means for each distribution pattern.

Description

Technical Field

[0001] Embodiments of the present invention relate to an information processing apparatus and an information processing program.

Background Art

[0002] Various techniques have been proposed for predicting the effects of coupon distribution to customers. However, those techniques do not provide information for determining the budget regarding coupon distribution. Under such circumstances, it has been desired to be able to support planning coupon distribution involving budget determination.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The problem to be solved by the present invention is to provide an information processing apparatus and an information processing program that can support planning coupon distribution involving budget determination.

Means for Solving the Problems

[0005] The information processing device of the embodiment comprises a calculation means, a selection means, a control means, and a first generation means. The calculation means calculates an effect amount, which is an indicator of the effectiveness of each of a plurality of distribution patterns for distributing coupons to a plurality of customers within a budget. The selection means selects one of the plurality of distribution patterns based on the effect amount calculated by the calculation means. The control means controls the calculation means and the selection means to perform the calculation of expected profits and the selection of distribution patterns for different budgets, respectively. The first generation means generates information that compares the magnitude of the effect amounts calculated by the calculation means for each of the plurality of distribution patterns selected by the selection means under the control of the control means. [Brief explanation of the drawing]

[0006] [Figure 1] A block diagram showing the main circuit configuration of an information processing device according to one embodiment. [Figure 2] Flowchart of the analysis process. [Figure 3] A subflow diagram related to the effect size calculation process. [Figure 4] A diagram illustrating an example of product data. [Figure 5] A diagram illustrating an example of coupon data. [Figure 6] A diagram illustrating an example of the calculation result for purchase probability. [Figure 7] A diagram illustrating an example of the calculation results for the probability of use. [Figure 8] A diagram illustrating an example of the result of effect size calculation. [Figure 9] A subflow in the first embodiment of the budget decision process. [Figure 10] Figure 8 shows the total effect size and total cost for each distribution pattern when the calculation results are obtained through the effect size calculation process. [Figure 11] A diagram showing an example of the first proposal screen. [Figure 12] A subflow in the second embodiment of the budget decision process. [Figure 13] A diagram illustrating a specific example of how the applicable budget is determined during the third budget approval process. [Figure 14] A diagram illustrating a specific example of how the applicable budget is determined during the fourth budget approval process. [Figure 15] Flowchart of the analysis process in the third embodiment. [Figure 16] A diagram showing an example of the second proposal screen. [Figure 17] Flowchart of the analysis process in the fourth embodiment. [Figure 18] Flowchart of the analysis process in the fifth embodiment. [Figure 19] A subflow in the fifth embodiment of the budget decision process. [Modes for carrying out the invention]

[0007] An example of an embodiment will be described below with reference to the drawings. Figure 1 is a block diagram showing the main circuit configuration of the information processing device 1 according to this embodiment. The information processing device 1 includes a processor 11, a main storage unit 12, a sub-storage unit 13, an input unit 14, a display unit 15, a communication unit 16, and a transmission line 17, etc.

[0008] A computer is configured to perform various information processing tasks by connecting the processor 11, the main memory unit 12, and the sub-memory unit 13 with a transmission line 17. The processor 11 corresponds to the central part of the computer described above. The processor 11 performs information processing according to information processing programs such as the operating system and application programs.

[0009] The main memory unit 12 corresponds to the main memory part of the computer. The main memory unit 12 includes a read-only memory area and a rewritable memory area. The main memory unit 12 stores part of the above information processing program in the read-only memory area. Also, the main memory unit 12 may store data necessary for the processor 11 to execute processes for controlling each part in the read-only memory area or the rewritable memory area. The main memory unit 12 uses the rewritable memory area as a work area by the processor 11.

[0010] The sub-memory unit 13 corresponds to the auxiliary memory part of the computer. The sub-memory unit 13 can utilize, for example, an EEPROM (electric erasable programmable read-only memory), an HDD (hard disc drive), an SSD (solid state drive), or other various well-known memory devices. The sub-memory unit 13 stores data used by the processor 11 to perform various processes and data generated by the processes in the processor 11. The sub-memory unit 13 may store the above information processing program. In this embodiment, the sub-memory unit 13 stores an analysis program PRA which is one of the information processing programs. A part of the storage area of the sub-memory unit 13 is used as an area for storing a customer database DBA, a product database DBB, and a coupon database DBC.

[0011] The input unit 14 inputs various instructions by an operator. As the input unit 14, well-known input devices such as a touch sensor, a key switch, and a keyboard can be used alone or in combination. The input device included in the input unit 14 is, for example, a touch sensor provided in a touch panel.

[0012] The display unit 15 performs various display operations to notify the operator of various information. The display unit 15 can use well-known display devices, such as liquid crystal displays and light-emitting devices such as LED lamps, either individually or in combination. One of the display devices included in the display unit 15 is, for example, a display device provided on a touch panel.

[0013] The communication unit 16 performs communication processing for data communication via the communication network 2. The communication unit 16 can use existing communication devices that comply with the communication method of the communication network 2. The communication network 2 may include the internet, VPN (virtual private network), LAN (local area network), public communication network, mobile communication network, etc., either alone or in appropriate combinations. The transmission path 17 includes an address bus, a data bus, and control signal lines, and transmits data and control signals exchanged between the connected parts.

[0014] The basic hardware of the information processing device 1 is expected to be, for example, a general-purpose personal computer. The transfer of the information processing device 1 is carried out, for example, with the analysis program PRA stored in the sub-storage unit 13. However, the hardware and the analysis program PRA may be transferred separately, either without the analysis program PRA stored in the sub-storage unit 13, or with a different version of the same type of application program stored in the sub-storage unit 13. The information processing device 1 may also be configured by writing the analysis program PRA to the sub-storage unit 13 in response to the operation of any operator. The transfer of the analysis program PRA can be carried out by recording it on a removable recording medium such as a magnetic disk, magneto-optical disk, optical disk, or semiconductor memory, or by communication over a network.

[0015] Furthermore, other devices, such as a server computer, can be used as the basic hardware for the information processing device 1. In cases where a server computer is used as the basic hardware, at least one of the input unit 14 and the display unit 15 may be externally attached rather than included in the information processing device 1. Alternatively, an input unit and a display unit provided in a client terminal that can communicate with the information processing device 1 via the communication network 2 may be used instead of the input unit 14 and the display unit 15.

[0016] Next, several embodiments of the operation of the information processing device 1 configured as described above will be explained. Note that the content of the processes described below is just an example, and it is possible to change the order of some processes, omit some processes, or add other processes as appropriate. For example, in the following explanation, some processes have been omitted in order to clearly explain the characteristic operation of this embodiment. For example, after transitioning from one processing state to another, the system may return to the processing state before the transition in response to instructions from the operator, but such processes have been omitted from the description. Or, for example, if some error occurs, processing to deal with that error may be performed, but such processes have been omitted from the description.

[0017] [First Embodiment] Information processing device 1 is used, for example, by a person in charge of planning sales promotions by mass retailers, such as distributing coupons to customers. Information processing device 1 then performs information processing to analyze the effects of coupon distribution and supports the work of the person in charge by presenting the analysis results. In other words, as an example, the person in charge in question becomes the operator of information processing device 1.

[0018] Prior to starting the information processing described later by the information processing device 1, the customer database DBA, the product database DBB, and the coupon database DBC are stored in the sub-storage unit 13. The customer database DBA is a database for managing customers who may be eligible to receive coupons, as well as for managing the transaction history of those customers. The product database DBB is a database for managing products that are the subject of transactions between mass retailers and customers. The coupon database DBC is a database for managing coupons to be distributed to customers.

[0019] If a predetermined operation to instruct the start of the analysis process is performed, for example, in the input unit 14, the processor 11 starts the analysis process based on the analysis program PRA. Figure 2 is a flowchart of the analysis process. As ACT1, the processor 11 acquires processing conditions related to the analysis process. In this embodiment, the processor 11 acquires the lower and upper limits of the budget and the number of iterations as processing conditions. For example, the processor 11 displays a GUI (graphical user interface) screen for specifying the processing conditions on the display unit 15, and acquires the processing conditions that the operator inputs by operating the input unit 14 accordingly. Alternatively, the processor 11 may read a setting file that represents the processing conditions and is pre-stored in the sub-storage unit 13. Alternatively, the processor 11 may acquire a similar setting file from any information processing device via the communication network 2.

[0020] As ACT2, processor 11 performs effect size calculation processing. Effect size calculation processing is the process of calculating the effect size, which is an indicator of the effect of distributing coupons managed by coupon database DBC to all customers managed by customer database DBA. Processor 11 may calculate the effect size for only a portion of the customers managed by customer database DBA. For example, processor 11 may narrow down the customers managed by customer database DBA to those that meet predetermined filtering conditions and calculate the effect size for those customers. Processor 11 may also calculate the effect size for only a portion of the coupons managed by coupon database DBC. For example, processor 11 may narrow down the coupons managed by coupon database DBC to those with a predetermined distribution period and calculate the effect size for those coupons.

[0021] Figure 3 shows a subflow of the effect size calculation process. As ACT211, processor 11 acquires various types of data, such as customer data, product data, and coupon data. For example, processor 11 reads customer data, product data, and coupon data from customer database DBA, product database DBB, and coupon database DBC. However, processor 11 may acquire some or all of the data by communicating with other information processing devices via communication network 2. For example, processor 11 may access a customer server that manages the customer database via communication network 2 and acquire customer data from this customer server. For example, processor 11 may access a product server that manages the product database via communication network 2 and acquire product data from this product server. For example, processor 11 may access a coupon server that manages the coupon database via communication network 2 and acquire coupon data from this coupon server.

[0022] Figure 4 shows an example of product data. The product data shown in Figure 4 includes, for each product, a product code to identify the product, a selling price, and a cost price. Note that Figure 4 shows data for only two products for simplification; in most cases, it includes data for more products. Furthermore, the product data may include additional information for each product, such as the product name.

[0023] Figure 5 shows an example of coupon data. The coupon data shown in Figure 5 includes a coupon code and discount amount for each coupon. Note that the coupon data in Figure 5 is an example where only one coupon is to be distributed; however, if multiple coupons are to be distributed, the discount amount should be included for each of those coupons. Furthermore, the coupon data may also include other information for each coupon, such as the coupon name. In this embodiment, the coupons to be analyzed are those that offer a discount service as a benefit. However, coupons that offer a discount on the product price may also be included in the analysis. In the case of such coupons being included in the analysis, the coupon data will include a discount rate instead of the discount amount.

[0024] In Figure 3, the processor 11, as ACT212, acquires various parameters of the machine learning model. For example, the processor 11 reads from the sub-storage unit 13 parameters of a trained machine learning model that predicts the probability of a customer purchasing each product based on whether or not they have a coupon. For example, the processor 11 reads from the sub-storage unit 13 parameters of a trained machine learning model that predicts the probability of a customer using each coupon. The processor 11 may acquire some or all of the various parameters by communicating with other information processing devices via the communication network 2.

[0025] As ACT213, processor 11 calculates the purchase probability for each product and the usage probability for each coupon for all customers using a predictive model. Figure 6 shows an example of the calculation results for the probability of purchase. Figure 7 shows an example of the calculation results for the probability of use.

[0026] In Figure 3, processor 11, designated as ACT214, calculates the expected profit and cost from the purchase probability, usage probability, product data, and coupon data. As ACT215, processor 11 calculates the effect size of the coupon for each customer. For example, processor 11 calculates the expected value of the profit increase due to the coupon for each customer, and the value obtained from this calculation is taken as the effect size. Also, for example, processor 11 calculates the expected value of the cost incurred by issuing the coupon to the customer, and the value obtained from this calculation is taken as the cost. Then processor 11 terminates the effect size calculation process. Thus, the expected value of profit increase due to coupons is just one example of an effect size. Other numerical values ​​that indicate the effect of coupon distribution, such as the expected value of increased store visits, can also be used as effect sizes. Figure 8 shows an example of the effect size calculation results.

[0027] In Figure 2, processor 11, as ACT3, performs budget determination processing. Budget determination processing is the process of determining the next budget to be applied to the optimization process described later (hereinafter referred to as the applied budget). Figure 9 shows a subflow in the first embodiment of the budget decision process.

[0028] As ACT311, processor 11 checks whether this is the first time the optimization process using the budget to be determined is being performed. For example, if the process has progressed from ACT2 to ACT3 in Figure 2, processor 11 determines that it is the first time and YES, and proceeds to ACT312. As ACT312, processor 11 checks the processing conditions that have been acquired to be applied to the current analysis process. In other words, processor 11 reads the processing conditions that were acquired in ACT1 in Figure 2 and stored in sub-storage unit 13.

[0029] In Figure 9, as ACT313, processor 11 sets multiple candidate budgets (hereinafter referred to as candidate budgets) according to the processing conditions confirmed above. For example, processor 11 sets the lower limit and upper limit included in the processing conditions as candidate budgets. Then, if the number of iterations included in the processing conditions is [3] or more, processor 11 selects [number of iterations - 2] values ​​from the values ​​between the lower limit and the upper limit to be the candidate budgets. For example, processor 11 sets each candidate budget such that the amount obtained by dividing the difference between the upper limit and the lower limit by [number of iterations - 1] is the difference between adjacent candidate budgets. As an example, if the lower limit is 0 yen, the upper limit is 300 yen, and the number of iterations is 4, processor 11 sets 0 yen, 100 yen, 200 yen, and 300 yen as candidate budgets. Thus, by processor 11 executing information processing based on the analysis program PRA, the computer with processor 11 as its central part functions as the first decision-making means.

[0030] Once processor 11 has finished setting all candidate budgets, it proceeds to ACT314. Processor 11 repeats the budget determination process shown in Figure 9, as will be described later, but for the second time and beyond, it determines NO at ACT311, skips ACT312 and ACT313, and proceeds to ACT314. As ACT314, processor 11 selects one of the candidate budgets that was not yet selected in ACT313 during the initial setup, as described above, and sets it as the applicable budget. With this, processor 11 completes the budget determination process and proceeds to ACT4.

[0031] As ACT4, processor 11 performs optimization processing. The optimization processing searches for a coupon distribution pattern (hereinafter referred to as the optimal pattern) that maximizes the total effect amount within the applied budget, based on the effect amount and cost for each customer recorded in ACT2, and records the found optimal pattern and the applied budget in, for example, sub-storage unit 13.

[0032] Figure 10 shows the total effect amount and total cost for each distribution pattern when the calculation results shown in Figure 8 are obtained in ACT2 in Figure 2. In the situation where the calculation results shown in Figure 8 are obtained through the fall rate calculation process, four distribution patterns are possible, as shown in Figure 10 as (a), (b), (c), and (d). (a) is a distribution pattern in which no coupon COA is distributed to either customer CUA or CUB, with an effect size of [0] and a cost of [0]. (b) is a distribution pattern in which coupon COAs are distributed only to customer CUAs, with an effect size of

[50] and a cost of

[25] . (c) is a distribution pattern in which coupon COAs are distributed only to customer CUBs, with an effect size of

[70] and a cost of

[30] . (d) is a distribution pattern in which coupon COAs are distributed to customers CUA and CUB respectively, with an effect size of

[0120] and a cost of

[55] .

[0033] In this way, processor 11 calculates the effect size for each of the multiple distribution patterns within the applicable budget. Thus, by having processor 11 perform information processing based on the analysis program PRA, the computer with processor 11 as its central component functions as a calculation means.

[0034] The processor 11 then determines the distribution pattern that yields the highest effect within the calculated budget as the optimal pattern. That is, when the budget is

[10] , distribution pattern (a) is determined to be the optimal pattern. When the budget is

[30] , distribution pattern (c) is determined to be the optimal pattern. When the budget is

[70] , distribution pattern (d) is determined to be the optimal pattern. However, in actual operation, the number of customers and coupons is large, and the number of distribution patterns is enormous, so it is not practical to calculate the effect size and cost for all distribution patterns as described above. Therefore, it is preferable to find the optimal pattern using a constrained optimization algorithm that can efficiently search for the optimal pattern within the applicable budget.

[0035] This optimization process is equivalent to selecting one of several distribution patterns based on the effect size, which is an example of expected profit. Thus, by having the processor 11 perform information processing based on the analysis program PRA, the computer with the processor 11 as its central component functions as a selection tool.

[0036] In ACT5, processor 11 checks whether the termination condition for the iteration of determining the optimal pattern described above has been met. For example, the termination condition is that ACT3 and ACT4 have been executed a number of times specified in the processing conditions obtained in ACT1. If the number of executions of ACT3 and ACT4 has not reached the specified number of iterations, processor 11 determines NO and returns to ACT3, repeating ACT3 and ACT4 as described above. In other words, processor 11 repeats the iterative steps, including the budget determination process and the optimization process, until the termination condition is met. However, at this time, processor 11 selects an unselected candidate budget as the applicable budget in ACT314 of ACT3, so ACT4 is repeated with a different applicable budget. Thus, by processor 11 executing information processing based on the analysis program PRA, the computer with processor 11 as its central component functions as a control means.

[0037] The termination conditions may be modified as appropriate, in accordance with the termination conditions of general optimization algorithms. For example, the termination condition could be set as when the effect size calculated in ACT4 is greater than or equal to a predetermined threshold. Alternatively, the termination condition could be set as when the difference between multiple effect sizes obtained from a certain number of optimization processes so far falls below a threshold, indicating a stagnation in the search. Furthermore, the processor 11 may simply confirm that any of the multiple termination conditions are met.

[0038] Then, when the number of iterations of the iterative step reaches the number of iterations included in the processing condition, processor 11 determines YES in ACT5 and proceeds to ACT6. As ACT6, the processor 11 creates a first proposal screen and displays the first proposal screen on the display unit 15. The first proposal screen is a screen that shows the optimal patterns for each candidate budget determined by the optimization process, making it possible to compare the magnitude of the effect of each optimal pattern.

[0039] Figure 11 shows an example of the first proposed screen. The first proposed screen shown in Figure 11 includes graph GRA and tables TAA, TAB, and TAC. Graph GRA is a graph in which the budget is shown on the horizontal axis and the effect size for each budget on the vertical axis. Table TAA shows the optimal pattern for a budget of "100 yen", the cost of that optimal pattern, and the effect size of that optimal pattern. Table TAB shows the optimal pattern for a budget of "200 yen", the cost of that optimal pattern, and the effect size of that optimal pattern. Table TAC shows the optimal pattern for a budget of "300 yen", the cost of that optimal pattern, and the effect size of that optimal pattern.

[0040] The graph GRA included in the first proposal screen is an example of information that shows the relative magnitudes of effect sizes for multiple optimal patterns in a comparable manner. Similarly, the tables TAA to TAC included in the first proposal screen are examples of information that shows the relative magnitudes of effect sizes for multiple optimal patterns in a comparable manner. Thus, by having the processor 11 execute information processing based on the analysis program PRA, the computer with the processor 11 as its central component functions as the first generation means.

[0041] As described above, the information processing device 1 of the first embodiment presents the first proposal screen to the operator. This allows the information processing device 1 to make the operator aware of what effects can be obtained with each budget and which distribution pattern with each budget is the most effective. Based on the insights from the first proposal screen, the operator can then plan the coupon distribution, including determining the budget. Furthermore, the information processing device 1 of the first embodiment determines multiple candidate budgets such that the difference in amounts between adjacent candidate budgets is equal. This allows the operator to intuitively understand the relationship between each candidate budget and its effect size.

[0042] [Second Embodiment] In the second embodiment, the processor 11 follows the same analysis processing flow as in the first embodiment as shown in Figure 2, but the budget decision processing in ACT3 is performed as follows. Figure 12 shows a subflow in the second embodiment of the budget decision process.

[0043] As ACT321, processor 11 checks the processing conditions that have been acquired to be applied to the current analysis process. In other words, processor 11 reads the processing conditions that were acquired in ACT1 in Figure 2 and stored in sub-storage unit 13. As ACT322, processor 11 checks whether iterative steps ACT3 and ACT4 in Figure 2 have been executed two or more times. For example, if processor 11 has progressed from ACT2 to ACT3 in Figure 2, or if it has returned to ACT3 for the first time from ACT5, it determines NO and proceeds to ACT323.

[0044] As ACT323, processor 11 determines the upper or lower limit included in the processing conditions as the applicable budget. For example, when processor 11 proceeds to ACT323 for the first time, it determines the upper limit as the applicable budget, and when it proceeds to ACT323 for the second time, it determines the lower limit as the applicable budget. With this, processor 11 completes ACT3 in Figure 2. On the other hand, if the processor 11 executes ACT4 in Figure 2 two or more times because the number of iterations included in the processing conditions is 3 or more, and then determines NO in ACT5 and returns to ACT3, it determines YES in ACT322 in Figure 12 and proceeds to ACT324.

[0045] As ACT324, processor 11 determines a point in a coordinate system with budget and effect size as two axes, with respect to the effect size calculated in the previously executed iteration step. In other words, processor 11 reads out each pair of budget and effect size recorded in the previously executed iteration step and determines the point in the above coordinate system for each pair. Since the iteration step has been executed more than twice, processor 11 will determine two or more points.

[0046] As ACT325, processor 11 calculates the slope for each line segment connecting adjacent points among the points determined by ACT324. As ACT326, processor 11 selects the maximum slope among those calculated by ACT325 and determines the amount corresponding to the midpoint of the line segment with the maximum slope as the applicable budget.

[0047] Figure 13 is a diagram illustrating a specific example of the determination of the applicable budget during the third budget approval process. Figure 13 shows the case where the upper limit of the processing conditions is 300 yen and the lower limit is 0 yen, and the pairs "Applicable budget = 300 yen: Effect size = PEA" and "Applicable budget = 0 yen: Effect size = 0" have already been recorded in the iterative step. In this case, points POA and POB are determined in ACT325, and the slope with respect to line segment LIA is calculated. However, since the line segment in question is only LIA, 150 yen is determined as the applicable budget, which corresponds to the midpoint MIA of this line segment LIA.

[0048] Figure 14 is a diagram illustrating a specific example of the determination of the applicable budget during the fourth budget approval process. Figure 14 shows the case where a third applicable budget is determined, as in the example in Figure 13, and an effect size called PEB is recorded for this applicable budget of 150 yen in the third iteration step. In this case, points POA, POB, and POC are determined in ACT325, and the slopes with respect to line segments LIB and LIC are calculated. Since the slope of line segment LIB is larger, 75 yen is determined as the applicable budget, corresponding to the midpoint MIB of line segment LIB.

[0049] Thus, by having the processor 11 execute information processing based on the analysis program PRA, the computer with the processor 11 as its central component functions as a second decision-making mechanism. Then, if processor 11 has determined the applicable budget as ACT326, it finishes the budget determination process.

[0050] In this way, the information processing device 1 of the second embodiment sequentially determines the applicable budget according to the execution results of the iterative steps. This allows the operator to recognize which budget will produce what effect and which distribution pattern with which budget is most effective, more effectively than in the first embodiment, regarding the appropriate budget.

[0051] [Third Embodiment] Figure 15 is a flowchart of the analysis process in the third embodiment. In Figure 15, the same reference numerals are used for processes identical to those in Figure 2, and their detailed explanations are omitted.

[0052] Processor 11 performs ACT1 to ACT4 in the same manner as in the first embodiment. After completing ACT4, it proceeds to ACT11. As ACT11, processor 11 calculates cost-effectiveness from the applied budget and effect size. Thus, by having processor 11 perform information processing based on the analysis program PRA, the computer with processor 11 as its central component functions as a determination means for determining cost-effectiveness.

[0053] The processor 11 then adds the calculated cost-benefit ratio to the set of optimal pattern and applied budget recorded in ACT4. That is, the processor 11 records the set of optimal pattern, applied budget, and cost-benefit ratio in, for example, the sub-storage unit 13. The processor 11 then proceeds to ACT5. Thus, in the fifth embodiment, the processor 11 includes the calculation of cost-benefit ratio in ACT11 in the iterative step.

[0054] Furthermore, if processor 11 determines YES in ACT5, it proceeds to ACT12. As ACT12, processor 11 creates a second proposal screen and displays this second proposal screen on display unit 15. The second proposal screen is a screen that allows comparison of the optimal pattern for each candidate budget determined in ACT4, the effect amount of that optimal pattern, and the cost-benefit ratio calculated in ACT11.

[0055] Figure 16 shows an example of the second proposed screen. The second proposed screen shown in Figure 16 includes graphs GRA and GRB, and tables TAD, TAE, and TAF. Graph GRA is identical to the one included in the first proposal screen. Graph GRB is a graph that shows the budget on the horizontal axis and the cost-effectiveness for each budget on the vertical axis. Table TAD is a table that adds the cost-effectiveness for a budget of "100 yen" to Table TAA. Table TAE is a table that adds the cost-effectiveness for a budget of "200 yen" to Table TAB. Table TAF is a table that adds the cost-effectiveness for a budget of "300 yen" to Table TAC.

[0056] The graph GRB included in the second proposal screen is an example of information that allows for the comparison of the cost-effectiveness of multiple optimal patterns. Similarly, the tables TAD to TAF included in the second proposal screen are examples of information that allows for the comparison of the cost-effectiveness of multiple optimal patterns. Thus, by having the processor 11 execute information processing based on the analysis program PRA, the computer with the processor 11 as its central component functions as a second generation means.

[0057] Thus, the information processing device 1 of the third embodiment presents the operator with a second proposal screen. This allows the information processing device 1 of the third embodiment to make the operator aware of what effects can be obtained with each budget, what the cost-effectiveness is, and which distribution pattern with which budget is the most effective. Based on the insights from the first proposal screen, the operator can then plan the coupon distribution, including determining the budget.

[0058] [Fourth Embodiment] Figure 17 is a flowchart of the analysis process in the fourth embodiment. In Figure 17, the same reference numerals are used for processes identical to those in Figure 2, and their detailed explanations are omitted. Processor 11 performs ACT1 to ACT4 in the same manner as in the first embodiment. After completing ACT4, it proceeds to ACT6. After completing ACT6, Processor 11 proceeds to ACT5. If Processor 11 determines YES in ACT5, it terminates the analysis process.

[0059] Thus, in the fourth embodiment, the processor 11 includes ACT6 in the iteration step. As a result, each time the processor 11 determines an optimal pattern, it creates and displays a first suggestion screen related to the optimal patterns determined so far. In other words, each time the information processing device 1 of the fourth embodiment finishes determining an optimal pattern for a budget, it sequentially updates the first suggestion screen so that the optimal patterns determined so far can be compared. This makes it possible for the operator to decide which budget to adopt from the first suggestion screen even before the completion condition is met.

[0060] Alternatively, the processor 11 may be configured to receive a termination command from the operator while the analysis process is running, and terminate the analysis process in response to the termination command. In this way, the operator can terminate the analysis process if they can decide on the budget to be adopted before the completion conditions are met, as described above.

[0061] [Fifth Embodiment] Figure 18 is a flowchart of the analysis process in the fifth embodiment. In Figure 18, the same reference numerals are used for processes identical to those in Figures 2 and 15, and their detailed explanations are omitted. Processor 11 performs ACT1 to ACT4 and ACT11 in the same manner as in the third embodiment. After completing ACT11, Processor 11 performs ACT12 and then proceeds to ACT5. Thus, in the fifth embodiment, Processor 11 also includes ACT12 in the iterative steps. Furthermore, Processor 11 performs the budget determination process in ACT3 as follows.

[0062] Figure 19 shows a subflow in the fifth embodiment of the budget decision process. As ACT331, processor 11 checks the processing conditions that have been acquired to be applied to the current analysis process. In other words, processor 11 reads the processing conditions that were acquired in ACT1 and stored in sub-storage unit 13. As ACT332, processor 11 retrieves the application budget and cost-benefit pairs that have been determined in the previously executed iteration steps. That is, for example, processor 11 reads the budget and cost-benefit pairs that have been recorded in the sub-storage unit 13 in the previously executed iteration steps.

[0063] As ACT333, processor 11 calculates a regression model that predicts the relationship between budget and cost-effectiveness based on the application budget and cost-effectiveness pair obtained in ACT332. As ACT334, processor 11 uses the regression model calculated in ACT333 to derive a budget that is likely to be cost-effective (hereinafter referred to as the superior budget). For example, a Gaussian process regression model can be applied to the regression model. ACT333 and ACT334 are optimization processes that predict the relationship between budget and cost-effectiveness calculated in ACT4 and ACT11 in Figure 18 from already calculated budget and cost-effectiveness pairs, and then search for an applicable budget with superior cost-effectiveness based on that prediction. As ACT335, processor 11 determines the preferred budget derived in ACT334 as the applicable budget. With this, processor 11 terminates the budget determination process.

[0064] If processor 11 determines YES in ACT5 in Figure 18, it proceeds to ACT21. As ACT21, processor 11 selects the optimal pattern that maximizes cost-effectiveness calculated in ACT11 by repeating the iterative steps. Thus, by having processor 11 perform information processing based on the analysis program PRA, the computer with processor 11 as its central component functions as a selection means.

[0065] As ACT22, processor 11 creates a third proposal screen that displays the selected optimal pattern and the budget associated with that optimal pattern. The third proposal screen is envisioned to be a screen that allows the operator to identify which of the two previously displayed second proposal screens is the optimal pattern selected as described above. Then, processor 11 waits for a condition to be met, such as when the operator gives an instruction to end the display, and then terminates the analysis process shown in Figure 18. In this way, the information processing device 1 of the fifth embodiment determines the budget and optimal pattern that maximize cost-effectiveness and presents them to the operator. This makes it easy for the operator to recognize the most cost-effective distribution pattern and the associated budget.

[0066] Each of the above embodiments can be modified in various ways as follows. For effect size, it is acceptable to calculate multiple different values, such as the expected value of increased profits and the expected value of increased customer visits.

[0067] The processor 11 may also make it possible to compare the magnitudes of two or more of the multiple numerical values ​​calculated as effect sizes on the first or second suggestion screen.

[0068] The processor 11 does not need to display either the graph GRA or the tables TAA to TAC on the first proposed screen.

[0069] Processor 11 may choose not to display the graph GRA on the second proposed screen. Furthermore, the second proposed screen may choose not to display either the graph GRB or the tables TAA-TAC.

[0070] The processor 11 may, instead of generating the first or second suggested screen, or in addition to generating the first or second suggested screen, generate information other than the screen, such as data that represents various information using only text and not images.

[0071] Instead of displaying the first or second suggestion screen on the display unit 15, the processor 11 may output screen data representing the first or second suggestion screen, or other information separate from the screen as described above, to another information processing device, for example, through communication via the communication network 2.

[0072] The information processing device 1 may be configured to support the planning of coupon distribution by food and beverage providers or businesses that provide services that do not involve the exchange of goods.

[0073] The selection rules in ACT21 in Figure 18 may be determined as appropriate by, for example, the person who determined the specifications for the information processing device 1, or the administrator of the information processing device 1. For example, the selection rule could be that the cost-effectiveness is above a predetermined threshold, and all optimal patterns that meet this selection rule could be selected.

[0074] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims of the invention and its equivalents. [Explanation of Symbols]

[0075] 1... Information processing device, 2... Communication network, 11... Processor, 12... Main memory unit, 13... Sub memory unit, 14... Input unit, 15... Display unit, 16... Communication unit, 17... Transmission line.

Claims

1. A calculation method for calculating an effect size, which is an indicator of the effectiveness of each of several distribution patterns for distributing coupons to multiple customers within a budget, A selection means that selects one of the plurality of distribution patterns based on the effect amount calculated by the calculation means, Control means for controlling the calculation means and the selection means so that the calculation of expected profits and the selection of distribution patterns are performed with respect to different budgets, A first generation means generates information that, under the control of the control means, represents a plurality of distribution patterns selected by the selection means in a manner that allows for comparison of the magnitude of the effect amounts calculated by the calculation means for each distribution pattern, An information processing device equipped with the following.

2. A determination means for determining cost-effectiveness for each of the multiple distribution patterns selected by the selection means under the control of the control means, based on the expected profit calculated by the calculation means and the budget applied when calculating the expected profit, A second generation means generates information that allows comparison of the cost-effectiveness of each of the multiple distribution patterns selected by the selection means under the control of the control means, as determined by the determination means. The information processing apparatus according to claim 1, further comprising:

3. A calculation method for calculating an effect size, which is an indicator of the effectiveness of each of several distribution patterns for distributing coupons to multiple customers within a budget, A selection means that selects one of the plurality of distribution patterns based on the effect amount calculated by the calculation means, Control means for controlling the calculation means and the selection means so that the calculation of expected profits and the selection of distribution patterns are performed with respect to different budgets, A determination means for determining cost-effectiveness for each of the multiple distribution patterns selected by the selection means under the control of the control means, based on the effect amount calculated by the calculation means and the budget applied when calculating that effect amount, A selection means that, based on the cost-effectiveness determined by the determination means, selects at least one distribution pattern from among a plurality of distribution patterns selected by the selection means under the control of the control means, An information processing device equipped with the following.

4. A first determination means that determines a plurality of budgets to be applied to calculate the effect size by the calculation means, based on the lower and upper limits of the budget and the number of times the effect size is calculated and the distribution pattern is selected. The information processing apparatus according to claim 1 or claim 3, further comprising:

5. A second determination means controls the calculation means and the selection means to sequentially perform the calculation of effect size and the selection of distribution patterns, and determines the budget to be applied to next calculate the effect size by the calculation means based on the distribution pattern selected by the selection means and the effect size calculated by the calculation means with respect to that distribution pattern. The information processing apparatus according to claim 1 or claim 3, further comprising:

6. The computer installed in the information processing device, A calculation method for calculating an effect size, which is an indicator of the effectiveness of each of several distribution patterns for distributing coupons to multiple customers within a budget, A selection means that selects one of the plurality of distribution patterns based on the effect amount calculated by the calculation means, Control means for controlling the calculation means and the selection means so that the calculation of expected profits and the selection of distribution patterns are performed with respect to different budgets, A first generation means generates information that, under the control of the control means, represents a plurality of distribution patterns selected by the selection means in a manner that allows for comparison of the magnitude of the effect amounts calculated by the calculation means for each distribution pattern, An information processing program that enables a function to work.

Images