Information processing device and program

Abstract

An information processing device and a program for efficiently conducting product sales activities in financial institutions are provided. [Solution] The information processing device comprises: a selection receiving unit that receives from a user a selection of a predictive model from among a plurality of predictive models for predicting a score indicating the probability of contract success, the predictive model being machine-learned using training data that can correlate with the success or failure of a contract for a product offered by a financial institution; a customer data receiving unit that receives input of customer data in which customer identification information and the plurality of variables are associated; an extraction unit that extracts one or more variables from the plurality of variables contained in the customer data that correspond to the predictive model selected by the selection receiving unit from the user; a generation unit that generates customer data to which the score has been assigned by inputting customer data that includes at least the one or more variables extracted by the extraction unit into the predictive model; and an output unit that outputs the customer data to which the score has been assigned.

Description

[Technical Field]

[0001] The embodiments disclosed in the present specification and drawings relate to an information processing device and a program. [Background technology]

[0002] Traditionally, financial institutions such as banks have handled a wide range of business operations, and therefore offered a wide range of products related to their operations. This has resulted in huge operating costs as a large number of personnel are required to conduct sales activities to customers for each product.

[0003] In order to improve the efficiency of these sales activities, for example, financial institutions may analyze customer data to identify customers who are likely to be highly effective in sales, and then conduct sales activities only with these customers. However, analyzing customer data requires specialized knowledge and time, which still results in huge sales costs. In addition, because financial institutions offer a wide range of products, it is difficult to analyze customer data for all products.

[0004] On the other hand, for example, as in Patent Document 1, there are cases where sales support in financial institutions is provided by inputting search criteria to extract customers, outputting a map of the customer's location, and accepting modifications to the location, thereby supporting sales activities themselves. The technology in Patent Document 1 requires input of search criteria for each product, making it difficult to efficiently conduct sales activities for a wide range of products. For this reason, it is desirable to enable financial institutions to efficiently conduct product sales activities while reducing sales costs. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Patent No. 5497852 Summary of the Invention [Problem to be solved by the invention]

[0006] An object of the present invention is to provide an information processing device and a program that enable financial institutions to efficiently carry out sales activities for products. [Means for solving the problem]

[0007] The information processing device according to the present invention comprises: a selection receiving unit that receives from a user a selection of one of a plurality of prediction models generated for each product offered by the financial institution, the prediction models being machine-learned using training data that is a set of data associating the success or failure of a contract for the product with a plurality of variables that may be correlated with the success or failure of the contract, for predicting a score indicating the success probability of the contract for a combination of the variables whose success or failure of the contract is unknown; a customer data receiving unit that receives input of customer data in which customer identification information for identifying a customer is associated with the plurality of variables; an extraction unit that extracts one or more variables corresponding to the prediction model selected by the selection receiving unit from the user, from among a plurality of variables included in the customer data received by the customer data receiving unit; a generation unit that generates customer data to which the score is assigned by inputting customer data including at least one or more variables extracted by the extraction unit into the prediction model; an output unit that outputs the customer data to which the score has been assigned; Equipped with. [Effects of the Invention]

[0008] According to the present invention, it is possible to provide an information processing device and a program that enable financial institutions to efficiently carry out sales activities for products. [Brief explanation of the drawings]

[0009] [Figure 1] FIG. 1 is a block diagram illustrating an example of a configuration of an information processing device according to an embodiment. [Figure 2]FIG. 10 is a diagram showing an example of customer data. [Figure 3] FIG. 10 is a diagram showing an example of types of variables and formats of variables included in customer data. [Figure 4] FIG. 10 is a diagram showing an example of training data for a savings asset formation model and a card loan model. [Figure 5] FIG. 10 is a diagram illustrating an example of processing content of an extraction unit. [Figure 6] 10 is a flowchart illustrating a score output process according to an embodiment. [Figure 7] FIG. 10 is a diagram illustrating an example of a home screen. [Figure 8] FIG. 10 is a diagram illustrating an example of a model selection screen. [Figure 9] FIG. 10 is a diagram showing an example of a pull-down menu displayed on a model selection screen. [Figure 10] FIG. 10 is a diagram showing an example of a model list output on a model selection screen. [Figure 11] FIG. 10 is a diagram illustrating an example of a processing execution screen. [Figure 12] FIG. 10 is a diagram illustrating an example of a screen displayed while a file is being read. [Figure 13] FIG. 10 is a diagram illustrating an example of a process execution screen on which customer data has been uploaded. [Figure 14] FIG. 10 is a diagram illustrating an example of customer data including a plurality of variables extracted by an extraction unit. [Figure 15] FIG. 10 is a diagram showing an example of a screen during file analysis. [Figure 16] FIG. 10 is a diagram illustrating an example of a generation completion screen. [Figure 17] FIG. 10 is a diagram illustrating an example of customer data to which scores are assigned. DETAILED DESCRIPTION OF THE INVENTION

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0011] An information processing device 1 according to this embodiment will be described with reference to FIG. 1. FIG. 1 is a block diagram showing an example of the configuration of the information processing device 1 according to an embodiment. The information processing device 1 is a device for processing customer data. This information processing device 1 is, for example, an on-premise type. Note that the information processing device 1 may also be a cloud type, and the cloud type information processing device 1 may provide the functions and processes described below in the form of, for example, SaaS (Software as a Service) or cloud computing. The information processing device 1 is used, for example, by a person in charge of a financial institution. The financial institution is, for example, a bank. Note that the financial institution is not limited to a bank and may be an insurance company, a securities company, or the like. As shown in FIG. 1, the information processing device 1 includes a control unit 10, a communication unit 20, an operation input unit 30, a display unit 40, and a storage unit 50.

[0012] The control unit 10 has a selection receiving unit 11, a customer data receiving unit 12, an extraction unit 13, a generation unit 14, a notification unit 15, an output unit 16, and a correction unit 17. Each unit will be described in detail later. In this embodiment, each unit of the control unit 10 is realized by the processor of the information processing device 1 executing a predetermined program. Note that at least one of the units of the control unit 10 may be configured by hardware. Furthermore, the functions of the control unit 10 may be realized by cooperation between multiple information processing devices that can communicate with each other.

[0013] The communication unit 20 is an interface for transmitting and receiving information to and from other information processing devices (such as a web server of a financial institution) via wireless or wired communication. The communication method and standard used by the communication unit 20 are not particularly limited.

[0014] The operation input unit 30 is an interface for the user of the information processing device 1 to input various information, and is composed of information input means such as a keyboard, a mouse, a touch panel, etc. The user uses the operation input unit 30 to, for example, select one prediction model from multiple prediction models or upload customer data.

[0015] The customer data D1 according to this embodiment will be described with reference to FIGS. 2 and 3. FIG. 2 is a diagram showing an example of customer data. FIG. 3 is a diagram showing an example of the types of multiple variables included in the customer data and the formats of the variables. As shown in FIG. 2, the customer data D1 is data in which a customer identification ID is associated with multiple variables. This customer identification ID is an example of customer identification information for identifying a customer. In the example shown in FIG. 2, the customer data D1 includes, for example, ten variables as multiple variables: a variable related to age, a variable related to gender, a variable related to the amount of salary deposit (salary deposit amount), a variable related to the amount of utility bills paid, a variable related to the number of outgoing remittances, a variable related to the number of incoming remittances, a variable related to the ending balance of fixed deposits, a variable related to the ending balance of liquid deposits, a variable related to the ending balance of a residential mortgage, and a variable related to whether or not the customer is registered for Internet banking (IB).

[0016] As shown in Figure 3, variables related to age are shown as age in the target year and month. Variables related to gender are shown as 1 for females and 0 for non-females. Variables related to salary deposit amount (salary deposit amount), utility bill payment amount, fixed deposit balance at the end of the year, liquid deposit balance at the end of the year, and residential mortgage balance at the end of the year are shown as amounts in the target year and month. Variables related to the number of outgoing remittances and the number of incoming remittances are shown as numbers in the target year and month. Variables related to whether or not someone is registered for internet banking are shown as 1 if they are registered and 0 if they are not. In other words, each of the multiple variables is shown as a value.

[0017] The number of variables included in the customer data is not limited to 10. That is, the number of variables included in the customer data D1 is arbitrary and may be between 2 and 9, or may be 11 or more. Furthermore, the variables included in the customer data are not limited to variables related to age, gender, salary deposit amount (salary deposit amount), utility bill payment amount, number of outgoing remittances, number of incoming remittances, end-of-year balance of fixed deposits, end-of-year balance of liquid deposits, end-of-year balance of residential mortgage, and whether or not the customer has registered for internet banking, but may also include other variables.

[0018] 2 includes a customer ID as customer identification information, but the customer identification information is not limited to a customer ID. In other words, the type of customer identification information is arbitrary, and the customer identification information may be, for example, a name, an address, etc.

[0019] The display unit 40 has a video display means such as a liquid crystal display or an organic EL display, and displays a selection screen for selecting a prediction model (described later), a screen for uploading customer data, customer data, etc.

[0020] The storage unit 50 is configured with a semiconductor memory and / or a hard disk drive. The storage unit 50 stores, for example, a program for processing executed by the control unit 10 and data required for the processing. The storage unit 50 also stores multiple prediction models. The multiple prediction models are generated for each product offered by the financial institution. Each of the multiple prediction models is machine-learned using training data, which is a collection of data associating the success or failure of a product contract with multiple variables that may correlate with the success or failure of the contract. The prediction models predict a score indicating the probability of contract success (ease of contracting) for a combination of variables for which the success or failure of the contract is unknown. Examples of products offered by financial institutions include savings products, mortgages, foreign currency deposits, card loans, and credit cards. The learning method for machine learning may be supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, or deep learning, or a combination of these learning methods. The storage unit 50 also stores extraction conditions used by the extraction unit 13 (described later) to extract variables corresponding to the prediction model from multiple variables included in customer data. The extraction conditions are associated with, for example, the name of a prediction model and the types of multiple variables to be input to the prediction model. As shown in Figure 1, a storage unit 50 according to this embodiment stores a plurality of prediction models, including a savings asset formation model 51, a foreign currency deposit account opening model 52, a credit card model 53, a mortgage model 54, and a card loan model 55.

[0021] The savings asset formation model 51 is a model for predicting a score indicating the probability of success in a savings product contract for a combination of variables whose success or failure in the contract is unknown, which has been machine-learned using training data, which is a collection of data that associates the success or failure of a savings product contract with a plurality of variables that may correlate with the success or failure of the savings product contract. In other words, by inputting a combination of unknown variables in customer data into this savings asset formation model 51, a score indicating the level of the customer's need for savings asset formation is predicted as a score indicating the success probability of the savings product contract.

[0022] The foreign currency deposit account opening model 52 is a model for predicting a score indicating the probability of success in a foreign currency deposit contract for a combination of variables whose success or failure in the foreign currency deposit contract is unknown, which has been machine-learned using training data, which is a collection of data associating the success or failure of the foreign currency deposit contract with a plurality of variables that may correlate with the success or failure of the foreign currency deposit contract. In other words, by inputting a combination of unknown variables in the customer data into this foreign currency deposit account opening model 52, a score indicating the high probability of the customer opening a foreign currency deposit account is predicted as a score indicating the success probability of the foreign currency deposit contract.

[0023] The credit card model 53 is a model for predicting a score indicating the probability of success in signing up for a credit card for a combination of variables whose success or failure in signing up for a credit card is unknown, which has been machine-learned using training data, which is a collection of data associating the success or failure of a credit card agreement with a plurality of variables that may be correlated with the success or failure of a credit card agreement. In other words, by inputting a combination of unknown variables in customer data into the credit card model 53, a score indicating the likelihood of the customer signing up for a credit card is predicted as a score indicating the success probability of signing up for a credit card.

[0024] The mortgage model 54 is a model for predicting a score indicating the probability of success in concluding a mortgage contract for a combination of variables whose success or failure is unknown, which has been machine-learned using training data, which is a collection of data associating the success or failure of a mortgage contract with multiple variables that may correlate with the success or failure of the mortgage contract. In other words, by inputting a combination of unknown variables in the customer data into this mortgage model 54, a score indicating the likelihood of the customer applying for a mortgage is predicted as a score indicating the success probability of concluding a mortgage contract.

[0025] The card loan model 55 is a model for predicting a score indicating the probability of success in a card loan contract for a combination of variables whose success or failure in the card loan contract is unknown, which has been machine-learned using training data, which is a collection of data associating the success or failure of a card loan contract with a plurality of variables that may be correlated with the success or failure of the card loan contract. In other words, by inputting a combination of unknown variables in customer data into this card loan model 55, a score indicating the likelihood of the customer signing up for a card loan is predicted as a score indicating the success probability of the card loan contract.

[0026] 1, five prediction models are stored in the storage unit 50 as the plurality of prediction models, but the number of prediction models stored in the storage unit 50 is not limited to five. That is, the number of prediction models stored in the storage unit 50 is arbitrary, and may be two to four, or six or more.

[0027] Furthermore, the types of prediction models stored in the storage unit 50 are not limited to the savings asset formation model 51, the foreign currency deposit account opening model 52, the credit card model 53, the mortgage model 54, and the card loan model 55. For example, the storage unit 50 may store an investment trust account opening model, etc.

[0028] FIG. 4 is a diagram showing an example of training data for each of the savings asset formation model 51 and the card loan model 55. FIG. 4(a) is a diagram showing an example of training data used in machine learning when generating the savings asset formation model 51. In the example shown in FIG. 4(a), the training data TD1 for the savings asset formation model 51 is a collection of data in which the success or failure of the contract for the savings asset formation model 51 and a plurality of variables that may be correlated with the success or failure of the contract for the savings asset formation model 51 are associated with a variable related to age, a variable related to gender, a variable related to the amount of salary deposited (salary deposit amount), a variable related to the amount of utility bills paid, a variable related to the end-of-term balance of fixed deposits, and a variable related to whether or not the investor is registered with an IB. Then, machine learning of the savings asset formation model 51 is performed using the success or failure of the contract in the training data TD1 as the objective variable, and the variables related to age, gender, salary deposit amount (salary deposit amount), utility bill payment amount, fixed deposit balance, and IB registration status as explanatory variables, to generate the savings asset formation model 51. Note that "1" in the contract success or failure indicates that the contract has been concluded, and "0" indicates that the contract has not been concluded.

[0029] 4(b) is a diagram showing an example of training data used in machine learning when generating the card loan model 55. In the example shown in FIG. 4(b), the training data TD2 for the card loan model 55 is a collection of data in which the success or failure of the contract for the card loan model 55 and a plurality of variables that may be correlated with the success or failure of the contract for the card loan model 55 are associated with a variable related to age, a variable related to gender, a variable related to the amount of salary deposited (salary deposit amount), a variable related to the amount of utility bills paid, a variable related to the number of outgoing remittances, a variable related to the end balance of fixed deposits, a variable related to the end balance of liquid deposits, and a variable related to the end balance of a home loan for a person's residence. Then, machine learning of card loan model 55 is performed using the success or failure of the contract in training data TD2 as the objective variable, and variables related to age, gender, salary transfer amount (salary transfer amount), utility bill payment amount, number of outgoing remittances, end-of-year balance of fixed deposits, end-of-year balance of liquid deposits, and end-of-year balance of home loan for residential use as explanatory variables, and card loan model 55 is generated.

[0030] Next, each part of the control unit 10 will be described in detail.

[0031] The selection receiving unit 11 receives a selection of one prediction model from a plurality of prediction models from a user. The customer data receiving unit 12 receives input of customer data D1. The types of the plurality of variables included in the customer data D1 received by the customer data receiving unit 12 are the same regardless of the plurality of prediction models.

[0032] The extraction unit 13 extracts one or more variables corresponding to the prediction model selected by the selection receiving unit 11 from among the multiple variables included in the customer data D1 received by the customer data receiving unit 12. Specifically, the extraction unit 13 extracts variables required as inputs for the prediction model from the multiple variables included in the uploaded customer data D1.

[0033] FIG. 5 is a diagram illustrating an example of the processing performed by the extraction unit 13. As shown in FIG. 5, the types of multiple variables included in the customer data D1 accepted by the customer data acceptance unit 12 are the same regardless of the multiple prediction models. Therefore, the customer data D1 is uploaded, including variables related to age, gender, the amount of salary transfer (salary transfer amount), the amount of utility bill payments, the number of outgoing remittances, the number of incoming remittances, the end-of-year balance of fixed deposits, the end-of-year balance of liquid deposits, the end-of-year balance of residential mortgage loans, and whether or not the customer has registered an IB. If the savings asset formation model 51 is selected as the prediction model, the extraction unit 13 extracts, from the multiple variables included in the customer data D1, variables related to age, gender, the amount of salary transfer (salary transfer amount), the amount of utility bill payments, the end-of-year balance of fixed deposits, and the whether or not the customer has registered an IB.

[0034] 5, the types of multiple variables included in the customer data D1 accepted by the customer data accepting unit 12 are the same regardless of the multiple prediction models, so customer data D1 is uploaded, including variables related to age, gender, salary deposit amount (salary deposit amount), utility bill payment amount, number of outgoing remittances, number of inbound remittances, fixed deposit balance, liquid deposit balance, residential mortgage balance, and whether or not the customer has registered an IB. If the card loan model 55 is selected as the prediction model, the extracting unit 13 extracts, from the multiple variables included in the customer data D1, variables related to age, gender, salary deposit amount (salary deposit amount), utility bill payment amount, number of outgoing remittances, fixed deposit balance, liquid deposit balance, and residential mortgage balance.

[0035] In other words, the one or more variables extracted by the extraction unit 13 include at least one of a variable related to age, a variable related to gender, a variable related to the amount of salary deposited, a variable related to the amount of utility bills paid, a variable related to the number of outgoing remittances, a variable related to the number of incoming remittances, a variable related to the ending balance of fixed deposits, a variable related to the ending balance of liquid deposits, a variable related to the ending balance of a home loan for a person's residence, and a variable related to whether or not the person is registered for internet banking.

[0036] The generation unit 14 generates customer data to which a score indicating the probability of success of the contract is assigned by inputting customer data including at least one or more variables extracted by the extraction unit 13 into a prediction model. The notification unit 15 notifies the user that the generation of the customer data to which a score is assigned has been completed.

[0037] The output unit 16 outputs the customer data to which the scores have been assigned. The correction unit 17 corrects the value of a variable to 0 when the multiple variables extracted by the extraction unit 13 include a variable with no value.

[0038] 6 is a flowchart illustrating a score output process according to an embodiment. This score output process outputs a model selection screen, accepts the selection of a prediction model, accepts file uploads, extracts multiple variables according to the model, generates scored customer data, and outputs the scored customer data. This score output process is executed when a selection to generate a marketing score is accepted on the home screen.

[0039] As shown in Fig. 6, first, the selection receiving unit 11 in the control unit 10 outputs a home screen (step S11). Specifically, the selection receiving unit 11 outputs the home screen to the display unit 40. Fig. 7 is a diagram showing an example of the home screen. As shown in Fig. 7, a marketing score generation button B1 is output in a selectable manner on the home screen SC1.

[0040] Next, as shown in Fig. 6, the selection receiving unit 11 determines whether or not a selection for marketing score generation has been received (step S13). Specifically, the selection receiving unit 11 determines whether or not a selection for marketing score generation has been received by determining whether or not selection of the marketing score generation button B1 output on the home screen SC1 shown in Fig. 7 has been received from the user via the operation input unit 30. Then, in step S13, if the selection for marketing score generation has not been received (step S13: No), the selection receiving unit 11 waits and repeats the processing of step S13 until a selection for marketing score generation is received.

[0041] On the other hand, if a selection to generate a marketing score is accepted (step S13: Yes), the selection accepting unit 11 outputs a model selection screen (step S15). Specifically, the selection accepting unit 11 outputs a model selection screen on the display unit 40 from which a prediction model can be selected. FIG. 8 is a diagram showing an example of the model selection screen. As shown in FIG. 8, a pull-down button B2 for outputting a major classification of the prediction model is output on the model selection screen SC2. This major classification is a service provided by a financial institution. In this embodiment, each of the multiple prediction models is associated with a major classification.

[0042] Next, as shown in Fig. 6, the selection receiving unit 11 determines whether or not selection of the pull-down button B2 has been received (step S17). Specifically, the selection receiving unit 11 determines whether or not selection of the pull-down button B2 output on the model selection screen SC2 shown in Fig. 8 has been received from the user via the operation input unit 30. Then, in step S17, if selection of the pull-down button B2 has not been received (step S17: No), the process of step S17 is repeated and the unit waits until selection of the pull-down button B2 is received.

[0043] On the other hand, if a selection to generate a marketing score is received in step S17 (step S17: Yes), the selection receiving unit 11 outputs the major categories (step S19). Specifically, the selection receiving unit 11 outputs the major categories by expanding a pull-down menu. FIG. 9 is a diagram showing an example of the pull-down menu PM output on the model selection screen SC2. As shown in FIG. 9, the pull-down menu PM includes major categories such as debit / credit, investment products, loans, life events, insurance, and others. The user selects a business (service) for which prediction is required from among the major categories output, debit / credit, investment products, loans, life events, insurance, and others, via the operation input unit 30.

[0044] Next, as shown in Fig. 6, the selection receiving unit 11 determines whether or not a selection of a major category has been received (step S21). Specifically, the selection receiving unit 11 determines whether or not a selection of a major category in the pull-down menu PM output on the model selection screen SC2 shown in Fig. 9, that is, a selection of a service provided by a financial institution in the pull-down menu PM output on the model selection screen SC2 shown in Fig. 9, has been received from the user via the operation input unit 30. Then, in step S21, if a selection of a major category has not been received (step S21: No), the process of step S21 is repeated and the unit waits until a selection of a major category is received.

[0045] On the other hand, if a selection of a major category is accepted in step S21 (step S21: Yes), the selection accepting unit 11 outputs a model list (step S23). Specifically, the selection accepting unit 11 outputs, to the model selection screen SC2, a model list that is a list of prediction models for products related to the services of the financial institution whose selection was accepted in step S21. In this way, the selection accepting unit 11 accepts, from the user, the selection of one prediction model from among multiple prediction models.

[0046] Fig. 10 is a diagram showing an example of a model list output on the model selection screen SC2. In the example shown in Fig. 10, the selection receiving unit 11 has received a selection of an investment product from the pull-down menu PM, and therefore the model list LT1 output on the model selection screen SC2 includes information on the savings asset formation model 51 and a selection button B3a for selecting the savings asset formation model 51, information on the foreign currency deposit account opening model 52 and a selection button B3b for selecting the foreign currency deposit account opening model 52, and information on the investment trust account opening model and a selection button B3c for selecting the investment trust account opening model.

[0047] In this embodiment, the investment trust account opening model in the model list LT1 is not stored in the storage unit 50. For this reason, for example, in comparison with information on the savings asset formation model 51 and a selection button B3a for selecting the savings asset formation model 51, and information on the foreign currency deposit account opening model 52 and a selection button B3b for selecting the foreign currency deposit account opening model 52, the information on the investment trust account opening model and the selection button B3c are output in a different color and are output in such a way that the selection button B3c cannot be selected.

[0048] Next, as shown in Fig. 6, the selection receiving unit 11 determines whether or not a model selection has been accepted (step S25). Specifically, the selection receiving unit 11 determines whether or not a selection of a model has been accepted by determining whether or not a selection of the selection button B3 included in the model list LT1 on the model selection screen SC2 shown in Fig. 10 and a selection of the model decision button B4 on the model selection screen SC2 shown in Fig. 10 have been accepted from the user via the operation input unit 30. Then, if a model selection has not been accepted in step S25 (step S25: No), the selection receiving unit 11 waits and repeats the process of step S25 until a model selection is accepted.

[0049] On the other hand, if the selection of a model is accepted in step S25 (step S25: Yes), the customer data accepting unit 12 outputs a processing execution screen (step S27). Specifically, the customer data accepting unit 12 outputs to the display unit 40 a processing execution screen for executing a process of assigning a prediction score to the customer data D1.

[0050] Fig. 11 is a diagram showing an example of a processing execution screen. In the example shown in Fig. 11, the processing execution screen SC3 includes information IF1 related to the target model, a reception window W1, information IF2 related to the uploaded file path, a list LT2 related to the data set of multiple variables, and a processing start button B5.

[0051] The information IF1 on the processing execution screen SC3 is information on the name of the prediction model selected via the model selection screen SC2. In the example shown in Fig. 11, the savings asset formation model 51 was selected via the model selection screen SC2, so the customer data receiving unit 12 outputs "savings asset formation."

[0052] The reception window W1 on the process execution screen SC3 accepts file uploads. Specifically, the user uploads the customer data D1 by dragging and dropping the customer data D1 into the reception window W1 via the operation input unit 30, or by selecting the file selection button B6 displayed in the reception window W1 and selecting the customer data D1 to be uploaded from the storage unit 50.

[0053] The information IF2 about the uploaded file is, for example, the file path of the uploaded customer data D1. When the customer data accepting unit 12 accepts a file upload via the acceptance window W1, the customer data accepting unit 12 outputs information about the uploaded file.

[0054] The customer data accepting unit 12 also outputs a list of types of multiple variables that must be included in the customer data D1 to the processing execution screen SC3. In the example shown in Fig. 11, the list LT2 relating to the data sets of multiple variables that must be included in the customer data D1 includes, as data sets that must be included in the customer data D1, age, gender, salary deposit amount, utility bill payment amount, number of outgoing remittances, number of incoming remittances, ending balance of fixed deposits, ending balance of liquid deposits, ending balance of residential mortgage, and whether or not the customer has registered for internet banking. In other words, in the example shown in Fig. 10, it is output that the customer data D1 must include 10 variables.

[0055] 6, the customer data accepting unit 12 determines whether a file upload has been accepted (step S29). Specifically, the customer data accepting unit 12 determines whether a file upload including customer data D1 has been accepted in the acceptance window W1. If a file upload has not been accepted in step S29 (step S29: No), the process of step S29 is repeated and the unit waits until a file upload is accepted.

[0056] On the other hand, if a file upload is accepted in step S29 (step S29: Yes), the customer data accepting unit 12 reads the file (step S31). Specifically, the customer data accepting unit 12 reads the file including the customer data D1 for which the file upload was accepted.

[0057] Fig. 12 is a diagram showing an example of a screen displayed while a file is being read. As shown in Fig. 12, while the file is being read, the customer data accepting unit 12 outputs to the display unit 40 a screen SC31 indicating that the file is being read, superimposed on the processing execution screen SC3.

[0058] 6, the customer data accepting unit 12 outputs a file preview (step S33). Specifically, the customer data accepting unit 12 outputs one or more pieces of customer identification information included in the accepted customer data D1 and multiple variables associated with the customer identification information.

[0059] FIG. 13 is a diagram showing an example of the processing execution screen SC3 on which the customer data D1 has been uploaded. As shown in FIG. 13, the customer data accepting unit 12 outputs a preview list LT3 to the processing execution screen SC3. In the example shown in FIG. 13, this preview list LT3 includes three pieces of customer identification information included in the customer data D1 and multiple variables associated with each of the three pieces of customer identification information. By displaying a file preview in this way, the user can confirm whether the file containing the intended customer data D1 has been uploaded.

[0060] 6, the customer data accepting unit 12 determines whether or not selection of the process start button B5 has been accepted (step S35). Specifically, the customer data accepting unit 12 determines whether or not selection of the process start button B5 output on the process execution screen SC3 has been accepted from the user via the operation input unit 30. Then, in step S35, if selection of the process start button B5 has not been accepted (step S35: No), the process of step S35 is repeated and the unit waits until selection of the process start button B5 is accepted.

[0061] On the other hand, if selection of the process start button B5 is accepted in step S35 (step S35: Yes), the extraction unit 13 in the control unit 10 extracts multiple variables according to the model (step S37). Specifically, the extraction unit 13 extracts one or multiple variables according to the prediction model selected in step S25 from multiple variables included in the customer data D1, based on the extraction conditions stored in the storage unit 50.

[0062] 14 is a diagram showing an example of customer data including multiple variables extracted by the extraction unit 13. As shown in FIG. 14, when the extraction unit 13 receives the selection of the savings asset formation model 51, the extraction unit 13 extracts, for example, a variable related to age, a variable related to gender, a variable related to the amount of salary deposited by bank transfer (salary deposit amount), a variable related to the amount of utility bills paid, a variable related to the end-of-year fixed deposit balance, and a variable related to whether or not the customer has registered an IB, from multiple variables included in the customer data D1, based on the extraction conditions of the savings asset formation model 51. Then, as shown in FIG. 14, the extraction unit 13 generates customer data D2 including the extracted variable related to age, variable related to gender, variable related to the amount of salary deposited by bank transfer (salary deposit amount), variable related to the amount of utility bills paid, variable related to the end-of-year fixed deposit balance, and variable related to whether or not the customer has registered an IB.

[0063] 6, the correction unit 17 in the control unit 10 determines whether or not there is a null value (step S39). Specifically, the correction unit 17 determines whether or not there is a null value by determining whether or not there is a variable with no value among the multiple variables extracted by the extraction unit 13. More specifically, when the savings asset formation model 51 is selected, for example, the correction unit 17 determines whether or not there is a null value for each of the variables related to age, gender, salary transfer amount (salary transfer amount), utility bill payment amount, fixed deposit balance, and IB registration status.

[0064] Then, in step S39, if there is a null value, that is, if the multiple variables extracted by the extraction unit 13 include a variable with no value (step S39: Yes), the null value is replaced (step S41). Specifically, the correction unit 17 performs correction to replace the value of the null variable with 0. By performing numerical processing in this way, errors can be avoided and / or a certain level of accuracy can be ensured.

[0065] On the other hand, if there is no null value in step S39, that is, if the multiple variables extracted by the extraction unit 13 do not include a variable with no value (step S39: No), or after step S41, the generation unit 14 in the control unit 10 generates customer data to which a score has been assigned (step S43). Specifically, the generation unit 14 generates customer data to which a score has been assigned by inputting one or more variables extracted by the extraction unit 13 into the prediction model selected in step S25. More specifically, the generation unit 14 inputs one or more variables extracted by the extraction unit 13 into the prediction model selected in step S25, thereby obtaining the score output from the prediction model, and generates customer data to which a score has been assigned by associating the score output from the prediction model with the customer data D1 whose file upload was accepted by the customer data accepting unit 12 in step S29.

[0066] In the above-described step S43, the prediction model is assumed to output only a score, but the output of the prediction model is not limited to a score. That is, the output of the prediction model is arbitrary, and the prediction model may output scored customer data D2 by assigning a score to the customer data D2. In this case, the generation unit 14 may generate scored customer data by associating the scored customer data D2 with the customer data D1 whose file upload was accepted by the customer data accepting unit 12 in step S29, or may generate the scored customer data D2 output from the prediction model as scored customer data.

[0067] 6, notification unit 15 in control unit 10 notifies a generation completion message (step S45). Specifically, notification unit 15 notifies by outputting the generation completion message on processing execution screen SC3.

[0068] Fig. 15 is a diagram showing an example of a screen during file analysis, and Fig. 16 is a diagram showing an example of a generation completion message screen. As shown in Fig. 15, when the customer data accepting unit 12 accepts selection of the process start button B5 in step S35, it superimposes a screen SC32 indicating that the file is being analyzed on the process execution screen SC3 during the execution of the processes of steps S37 to S43, and outputs the screen SC32 to the display unit 40. Then, as shown in Fig. 16, when the generation of the customer data to which a score has been assigned in step S43 is completed, the notification unit 15 superimposes a generation completion screen SC33 on the screen SC32 indicating that the file is being analyzed, and outputs the screen SC33 to the display unit 40. The generation completion screen SC33 shown in Fig. 16 includes a generation completion message M1 and an OK button B7 for inputting that the generation completion message M1 has been confirmed.

[0069] Next, as shown in FIG. 6, the notification unit 15 determines whether to output the scored customer data (step S47). Specifically, the notification unit 15 determines whether to output the scored customer data by determining whether the generation completion message M1 has been confirmed. More specifically, in the example shown in FIG. 16, the notification unit 15 determines whether the generation completion message M1 has been confirmed by determining whether the OK button B7 included in the generation completion screen SC33 has been pressed, thereby determining whether to output the scored customer data. Then, if the scored customer data is not to be output in step S47 (step S47: No), the notification unit 15 waits and repeats the process of step S47 until the scored customer data is output.

[0070] On the other hand, if the customer data to which the scores have been assigned is to be output in step S47 (step S47: Yes), the output unit 16 in the control unit 10 outputs the customer data to which the scores have been assigned (step S49). Specifically, the output unit 16 outputs the customer data to which the scores generated in step S43 have been assigned to the display unit 40.

[0071] Fig. 17 is a diagram showing an example of customer data to which a score has been assigned. In the example shown in Fig. 17, when the customer data D1 shown in Fig. 2 is uploaded as a file in step S29, the output unit 16 outputs customer data D3 to which a score has been assigned, which is generated by assigning a score to the customer data D1. As shown in Fig. 17, the customer data D3 is assigned a score indicating the probability of success.

[0072] By executing the process of step S49, the score output process is completed.

[0073] As described above, the information processing device 1 of this embodiment accepts the selection of one of multiple prediction models, accepts customer data D1 in which customer identification information is associated with multiple variables, extracts one or more variables corresponding to the one prediction model from the customer data D1, and inputs customer data containing at least the extracted one or more variables into the prediction model to generate scored customer data and output the scored customer data.Therefore, based on the output scored customer data, the user can prioritize approaching customers with high scores, reducing the number of phone calls to customers and the number of direct mail recipients, thereby enabling financial institutions to efficiently conduct product sales activities.

[0074] The above describes the embodiments of the present invention. Based on the above description, a person skilled in the art may be able to conceive additional effects and various modifications of the present invention, but the aspects of the present invention are not limited to the above-described embodiments. Various additions, modifications, and partial deletions are possible within the scope of the conceptual idea and spirit of the present invention, which can be derived from the content defined in the claims and their equivalents.

[0075] For example, each part of the control unit 10 may be realized by a processor of the information processing device 1 executing a predetermined program and using hardware resources to perform software processing, or may be realized by the implemented hardware itself. When configured by software, a program that realizes at least some of the functions of the control unit 10 may be stored on a recording medium such as a flexible disk or CD-ROM and read and executed by a computer. The recording medium is not limited to removable media such as magnetic disks and optical disks, but may also be fixed recording media such as a hard disk drive or memory.

[0076] In addition, a program that realizes at least a part of the functions of the control unit 10 may be distributed via a communication line (including wireless communication) such as the Internet. Furthermore, the program may be encrypted, modulated, or compressed and distributed via a wired line or wireless line such as the Internet, or stored on a recording medium. [Explanation of symbols]

[0077] 1...information processing device, 10...control unit, 11...selection reception unit, 12...customer data reception unit, 13...extraction unit, 14...generation unit, 15...notification unit, 16...output unit, 17...correction unit, 20...communication unit, 30...operation input unit, 40...display unit, 50...storage unit, 51...accumulation asset formation model, 52...foreign currency deposit account opening model, 53...credit card model, 54...mortgage housing loan model, 55...card loan model

Claims

1. a selection receiving unit that receives from a user a selection of one of a plurality of prediction models generated for each product offered by a financial institution, the prediction models being machine-learned using training data that is a set of data associating the success or failure of a contract for the product with a plurality of variables that may be correlated with the success or failure of the contract, for predicting a score indicating the success probability of the contract for a combination of variables whose success or failure of the contract is unknown; a customer data receiving unit that receives input of customer data in which customer identification information for identifying a customer is associated with the plurality of variables; an extraction unit that extracts one or more variables corresponding to the prediction model selected by the selection receiving unit from the user, from among a plurality of variables included in the customer data received by the customer data receiving unit; a generation unit that generates customer data to which the score is assigned by inputting customer data including at least one or more variables extracted by the extraction unit into the prediction model; an output unit that outputs the customer data to which the score has been assigned; An information processing device comprising:

2. The information processing device according to claim 1 , further comprising a storage unit that stores the plurality of prediction models.

3. The information processing device according to claim 1 , wherein types of the plurality of variables included in the customer data accepted by the customer data accepting unit are the same regardless of the plurality of prediction models.

4. The information processing apparatus according to claim 3 , wherein the customer data receiving unit outputs a list of types of the plurality of variables that need to be included in the customer data.

5. The information processing apparatus according to claim 1 , further comprising a correction unit that corrects a variable to a value of 0 when the plurality of variables extracted by the extraction unit includes a variable that has no value.

6. The information processing device according to claim 1 , wherein the customer data receiving unit outputs one or more pieces of customer identification information included in the customer data received by the customer data receiving unit and the plurality of variables associated with the customer identification information.

7. 2. The information processing device of claim 1, wherein the one or more variables extracted by the extraction unit include at least one of a variable related to age, a variable related to gender, a variable related to the amount of salary deposited, a variable related to the amount of utility bills paid, a variable related to the number of outgoing remittances, a variable related to the number of incoming remittances, a variable related to the ending balance of fixed deposits, a variable related to the ending balance of liquid deposits, a variable related to the ending balance of a residential mortgage, and a variable related to whether or not the person is registered for internet banking.

8. The information processing device according to claim 1 , further comprising a notification unit that notifies the user that generation of the customer data to which the score has been assigned has been completed.

9. 2. The information processing device according to claim 1, wherein the selection receiving unit outputs a model selection screen on which the prediction model can be selected, and when a selection of a service provided by the financial institution is received from the user on the model selection screen, outputs a model list on the model selection screen that is a list of prediction models for the product related to the service provided by the financial institution whose selection was received, thereby receiving a selection of one of the plurality of prediction models from the user.

10. The information processing device according to claim 1 , wherein the financial institution is a bank.

11. To the computer A process of accepting, from a user, the selection of one of a plurality of predictive models generated for each product offered by a financial institution, the plurality of predictive models being machine-learned using training data, which is a set of data associating the success or failure of a contract for the product with a plurality of variables that may be correlated with the success or failure of the contract, for predicting a score indicating the success probability of the contract for a combination of variables whose success or failure of the contract is unknown; a process of receiving input of customer data in which customer identification information for identifying a customer is associated with the plurality of variables; A process of extracting one or more variables corresponding to the prediction model selected by the user from among a plurality of variables included in the received customer data; generating customer data to which the score is assigned by inputting the extracted one or more variables into the prediction model; and outputting the customer data to which the score has been assigned.

Images