JP6840446B2 - Data processing device and data processing method - Google Patents

Description

The present invention relates to a data processing apparatus and a data processing method for processing customer purchase data.

In recent years, attention has been focused on analyzing customer purchasing behavior. Various types of large amounts of data (so-called big data) are useful for analyzing customer purchasing behavior. Patent Document 1 describes a system that generates one huge database by combining data having similar characteristics in a plurality of databases provided by different providers.

Patent No. 5649756

Patent Document 1 discloses a method of combining a plurality of databases to generate one huge database. On the other hand, it is not easy to analyze customer purchasing behavior based on a large amount of purchasing data contained in a huge database, and it is required to establish a method for analyzing customer purchasing behavior using a huge database. ..

The present invention has been made in view of the above points, and an object of the present invention is to output information for making it possible to analyze a customer's purchasing behavior based on the customer's purchasing data.

The data processing device according to the first aspect of the present invention includes a data acquisition unit that acquires customer attribute data indicating customer attributes in association with purchase data indicating information regarding the purchase of a product by the customer, and the data acquisition unit. Classify the customer using the index calculation unit that calculates an index indicating the purchase tendency of the customer for each customer based on the purchase data acquired by the data acquisition unit and the customer attribute data acquired by the data acquisition unit. A group generation unit that generates one or more groups including one or more customers, respectively, and statistics on the index of one or more customers belonging to each of the one or more groups are obtained by the one or more groups. Each group index has an output unit that outputs data in association with each of the one or more groups.

The index calculation unit calculates a plurality of different indexes for each customer based on the purchase data acquired by the data acquisition unit, and the output unit calculates a plurality of the statistics related to the plurality of indexes. It may be output as a plurality of different group indices.

The group generation unit may generate the group including one or more customers who are similar in combination of a plurality of different indices.

The data processing device may further have an estimation unit that estimates a part of the indexes by using an index other than the part of the indexes when a part of the indexes cannot be calculated. ..

The group generation unit may generate the group including one or more customers who share at least one of the customer's gender, age group, and place of residence included in the customer attribute data.

The index calculation unit may calculate the index based on the relationship between the sales data indicating information on the sale of the product by the store and the purchase data.

The index calculation unit may calculate the index by using the elapsed time from the release date of the product included in the sales data to the purchase date of the product included in the purchase data.

The index calculation unit may calculate the index by using the difference between the normal price of the product included in the sales data and the purchase price of the product included in the purchase data.

The index calculation unit extracts a specific attribute product having a specific attribute from the product based on the sales data, and calculates the index using the purchase amount of the specific attribute product included in the purchase data. You may.

The index calculation unit may calculate the index based on the relationship between the preset reference date and the purchase date of the product included in the purchase data.

The output unit may represent the group on a two-dimensional plane that represents the relationship between the group index and a value for the group that is different from the group index.

The data processing method of the second aspect of the present invention is performed from a storage unit that stores the customer attribute data indicating the customer attributes executed by the processor and the purchase data indicating the information regarding the purchase of the product by the customer in association with each other. A step of acquiring the customer attribute data and the purchase data, a step of calculating an index indicating the purchasing tendency of the customer for each customer based on the purchase data acquired by the acquisition step, and a step of acquiring the customer. By classifying the customers using the customer attribute data acquired by the customer, a step of generating one or more groups including one or more customers, and one or more groups belonging to each of the one or more groups. It has a step of outputting a statistic of a customer regarding the index as a group index of each of the one or more groups in association with each of the one or more groups.

The data processing device of the third aspect of the present invention is a data acquisition unit that acquires purchase data indicating information regarding the purchase of a product by a customer, and the purchase of the customer based on the purchase data acquired by the data acquisition unit. An index calculation unit that calculates a plurality of different indexes showing a tendency for each customer, a group generation unit that generates one or more groups including one or more customers whose combinations of different indexes are similar, and the above. A plurality of statistics on a plurality of the indexes of one or more customers belonging to each of the one or more groups are output as a plurality of different group indexes of the one or more groups in association with each of the one or more groups. It has an output unit.

According to the present invention, there is an effect that information for enabling analysis of a customer's purchasing behavior can be output based on the customer's purchasing data.

It is a schematic diagram of the data processing system which concerns on embodiment. It is a block diagram of the data processing system which concerns on embodiment. It is a schematic diagram of purchase data, product data and customer attribute data. It is a figure which shows the flowchart of the whole data processing method executed by the data processing system which concerns on embodiment. It is a figure which shows the flowchart of the exponential calculation process executed by the data processing system which concerns on embodiment. It is a schematic diagram for explaining an early adopter index, a price sensitive index, a limited product purchase index, a health-related product purchase index, and a premium product purchase index. It is a schematic diagram for demonstrating the preferential day purchase index and the payday after-payment purchase index. It is a schematic diagram of the exponential estimation processing by the exponential estimation part. It is a figure which shows the flowchart of the group index output processing executed by the data processing system which concerns on embodiment. It is a schematic diagram of the method in which a group generation part generates a group. It is a front view of the information terminal which displays the index list screen. It is a front view of the information terminal which displays the exponential distribution screen.

[Overview of data processing system S]
FIG. 1 is a schematic diagram of a data processing system S (information processing system) according to the present embodiment. The data processing system S includes a data processing device 1 (information processing device) and an information terminal 2. The data processing system S may include other devices such as servers and terminals.

The data processing device 1 is a computer that outputs information for making it possible to analyze a customer's purchasing behavior based on the customer's purchasing data. The data processing device 1 can communicate with the information terminal 2 via a network such as the Internet or a local area network.

The information terminal 2 is a computer that displays information output by the data processing device 1. The information terminal 2 is, for example, a mobile terminal such as a smartphone or a tablet terminal, or a personal computer. The information terminal 2 has a display unit such as a liquid crystal display for displaying information.

The outline of the process in which the data processing system S according to the present embodiment outputs information for enabling the analysis of the customer's purchasing behavior based on the purchasing data will be described below. The data processing device 1 stores in advance customer attribute data indicating the attributes of the customer and purchase data indicating information regarding the purchase of the product by the customer in the storage unit. Goods include goods and services purchased by customers. The customer attribute data includes, for example, the customer's gender, age, place of residence, and the like. The purchase data includes, for example, information that can identify the product (product ID), a purchase date, a purchase price, and the like. The data processing device 1 acquires customer attribute data and purchase data from one or more providers and stores them in a storage unit.

The data processing device 1 calculates an index indicating a customer's purchasing tendency for each customer based on the purchasing data stored in the storage unit. The index is at least one of the early adopter index, price sensitive index, limited product purchase index, health-related product purchase index, premium product purchase index, preferential day purchase index, and payday purchase index, which will be described later, and the customer purchases the product. It is a value that is an indicator of purchasing tendency.

Further, the data processing device 1 classifies the customers by using the customer attribute data stored in the storage unit, or by classifying the customers based on the combination of a plurality of indexes calculated for each customer. Generate one or more groups that contain more than one customer. Then, the data processing device 1 calculates a statistic regarding the index of one or more customers belonging to the group as the group index.

The data processing device 1 transmits information for displaying the calculated group index in association with the group to the information terminal 2. The data processing device 1 is not limited to this, and may output the group index by printing on paper using a printer, storing in a storage medium, or transmitting to the outside by communication.

As described above, the data processing system S according to the present embodiment outputs a group index in which the index indicating the customer's purchasing tendency is statistic for each customer group. Therefore, the user (analyst) who uses the data processing system S can grasp the purchasing tendency for each group of customers by referring to the group index, and can efficiently analyze the purchasing behavior of the customers.

[Configuration of data processing system S]
FIG. 2 is a block diagram of the data processing system S according to the present embodiment. In FIG. 2, the arrows indicate the main data flows, and there may be data flows other than those shown in FIG. In FIG. 2, each block shows not a hardware (device) unit configuration but a functional unit configuration. Therefore, the block shown in FIG. 2 may be mounted in a single device, or may be mounted separately in a plurality of devices. Data can be exchanged between blocks via any means such as a data bus, a network, or a portable storage medium.

The data processing device 1 has a control unit 11 and a storage unit 12. The control unit 11 includes a data acquisition unit 111, an index calculation unit 112, an index estimation unit 113, a group generation unit 114, a group index calculation unit 115, and an output unit 116. The storage unit 12 includes a purchase data storage unit 121, a product data storage unit 122, a customer attribute data storage unit 123, and a calculation result storage unit 124.

The storage unit 12 is a storage medium including a ROM (Read Only Memory), a RAM (Random Access Memory), a hard disk drive, and the like. The storage unit 12 stores in advance the program executed by the control unit 11. The storage unit 12 may be provided outside the data processing device 1, and in that case, data may be exchanged with the control unit 11 via a network.

The purchase data storage unit 121 stores purchase data. The product data storage unit 122 stores the product data. The customer attribute data storage unit 123 stores customer attribute data. The calculation result storage unit 124 stores the calculation results of the index and the group index by the data processing device 1. The purchase data storage unit 121, the product data storage unit 122, the customer attribute data storage unit 123, and the calculation result storage unit 124 may each be a storage area on the storage unit 12, or a database configured on the storage unit 12. It may be.

The control unit 11 is, for example, a processor such as a CPU (Central Processing Unit), and by executing a program stored in the storage unit 12, the data acquisition unit 111, the index calculation unit 112, the index estimation unit 113, and the group generation unit It functions as 114, a group index calculation unit 115, and an output unit 116. At least a part of the function of the control unit 11 may be executed by an electric circuit. Further, at least a part of the functions of the control unit 11 may be executed by a program executed via the network.

The data processing system S according to the present embodiment is not limited to the specific configuration shown in FIG. The data processing device 1 and the information terminal 2 may be configured by connecting two or more physically separated devices, respectively, by wire or wirelessly.

[Explanation of data processing method]
Hereinafter, the data processing method executed by the data processing system S according to the present embodiment will be described in detail. In the data processing device 1, the purchase data storage unit 121 stores in advance purchase data indicating information regarding the purchase of goods by the customer. The product data storage unit 122 stores in advance product data indicating information about the product. The customer attribute data storage unit 123 stores in advance customer attribute data indicating customer attributes.

FIG. 3A is a schematic diagram of the purchase data D1 stored in the purchase data storage unit 121. FIG. 3B is a schematic diagram of the product data D2 stored in the product data storage unit 122. FIG. 3C is a schematic diagram of the customer attribute data D3 stored in the customer attribute data storage unit 123.

The purchase data D1 includes a purchase ID (purchase identification information), a customer ID (customer identification information), a product ID (product identification information), a purchase date, and a purchase amount associated with each other. The purchase data D1 is, for example, receipt information recorded at a store, credit card payment information recorded by a credit card business operator at the time of payment of a product price, or electronic payment information recorded by an electronic payment business operator at the time of payment of a product price. is there. The purchase data D1 may further include a store ID (store identification information) for identifying the store where the customer made a purchase. In the present specification, the store includes both a physical store that sells products in the real world and a virtual store such as an EC (Electronic Commerce) site.

The purchase ID is identification information for identifying one purchase. The purchase ID may be automatically assigned to the purchase by the data processing device 1 or may be assigned to the purchase by the data provider. The customer ID is identification information for identifying the customer who made the purchase, for example, a customer membership number or an identification number automatically assigned to the customer.

The product ID is identification information for identifying a product purchased by a customer, and is, for example, a product code assigned to each product or an item code indicating an item (type) of the product. A product is, for example, an article purchased by a customer at a retail store, or a service (plan) contracted by a customer at a sports club, a travel agency, an insurance company, or the like. In the purchase data D1, a plurality of product IDs may be associated with one purchase ID.

The purchase date is the date on which the customer purchased the item. The purchase date may be the date on which the contract for the product is made. The purchase date may include the time as well as the date. The purchase price is the price paid by the customer to purchase the goods. The purchase price may be the price of each product or the total price of a plurality of products.

The product data D2 includes a product ID, a release date, a sold-out date, a normal price, and a product description associated with each other. The product ID is the same as the product ID of the purchase data D1. The release date is the date on which the product was released. The end-of-sale date is the date when the sale of the product ends. The normal price is a normal price (for example, a fixed price) when a product is sold. The normal price of the product data D2 may be different from the purchase price of the purchase data D1 due to a discount or the like. The product description is a character string indicating the attributes of the product. The product description may be the name of the product or the description of the product.

The customer attribute data D3 includes a customer ID, a gender, an age, and a place of residence. The customer ID is the same as the customer ID of the purchase data D1. Gender is the gender of the customer. Age is the age of the customer. For the age of the customer attribute data D3, an age group may be set instead of the age. The place of residence is the place where the customer resides. The particle size of the place of residence is arbitrarily determined and may include prefectures, towns and streets, or house numbers. The customer attribute data D3 may be information registered by the customer himself / herself, or may be information estimated from the appearance of the customer or the location of the store.

In FIGS. 3A to 3C, the purchase data D1, the product data D2, and the customer attribute data D3 are represented by a character string table for visibility, but each data is recorded in an arbitrary format. It may be, for example, any of character string data, numerical data and binary data. The purchase data D1, the product data D2, and the customer attribute data D3 may be recorded as a database, or may be recorded as a list listing the data.

The data included in the purchase data D1, the product data D2, and the customer attribute data D3 shown in FIGS. 3 (a) to 3 (c) is an example, and may not include some information or may include other information. Good. Two or more of the purchase data D1, the product data D2, and the customer attribute data D3 may be combined in advance and stored in the storage unit 12 as one data. In that case, in the following, of the combined data, the part corresponding to the purchase data D1 is called the purchase data, the part corresponding to the product data D2 is called the product data, and the part corresponding to the customer attribute data D3 is called the customer attribute. Called data.

FIG. 4 is a diagram showing a flowchart of the entire data processing method executed by the data processing system S according to the present embodiment. The flowchart of FIG. 4 is started by, for example, performing a predetermined start operation in the data processing device 1. First, the data processing device 1 performs an index calculation process for calculating an index indicating a customer's purchasing tendency for each customer based on the purchase data (S1).

Next, the data processing device 1 performs a group index output process of calculating a statistic related to the index calculated by the index calculation process and outputting it as a group index for a group including one or more customers (S2). Hereinafter, the index calculation process and the group index output process will be described in detail with reference to FIGS. 5 to 12.

FIG. 5 is a diagram showing a flowchart of the index calculation process S1 executed by the data processing system S according to the present embodiment. The index calculation process S1 is performed for each of a plurality of customers whose purchase data is stored in the purchase data storage unit 121. First, the data acquisition unit 111 acquires one or more purchase data associated with the customer ID of one customer whose index is to be calculated from the purchase data storage unit 121 (S11). Further, the data acquisition unit 111 acquires the customer attribute data associated with the customer ID from the customer attribute data storage unit 123. The data acquisition unit 111 may acquire only the purchase data including the purchase date within the period specified by the user.

Next, the data acquisition unit 111 specifies sales data indicating information regarding sales by the store of the product with the product ID included in the purchase data (S12). The sales data is information used for calculating each index described later, and includes a product release date, a sale date, a normal price, and a product attribute. In order to specify the sales data, the data acquisition unit 111 acquires the release date, the end date, the normal price, and the product description included in the product data associated with the product ID included in the purchase data in the product data storage unit 122. ..

When the product data does not include the release date, the data acquisition unit 111 determines the earliest purchase date as the release date among all the purchase data associated with one product ID in the purchase data storage unit 121. May be good. When the product data does not include the end date, the data acquisition unit 111 determines the latest purchase date as the end date among all the purchase data associated with one product ID in the purchase data storage unit 121. You may. When the product data does not include the normal price, the data acquisition unit 111 determines the mode of the purchase price as the normal price among all the purchase data associated with one product ID in the purchase data storage unit 121. You may.

In addition, the data acquisition unit 111 determines the product attribute indicating the product attribute of the product ID included in the purchase data. Specifically, the data acquisition unit 111 determines the product attribute of the product as a limited product when the time from the release date of the product to the end date is within a predetermined value (for example, within one year). The data acquisition unit 111 determines the product attribute of the product as a premium product when the normal price of the product is higher than the average price of the products of the same type by a predetermined ratio or more. In this case, the product data includes the product type (item). The data acquisition unit 111 specifies the product type of the product ID included in the purchase data from the product data, the purchase data storage unit 121 extracts all the purchase data of the specified type of product, and the extracted purchase data includes the product type. The average purchase price is calculated as the average price. The average price for each type of product may be stored in advance in the storage unit 12.

The data acquisition unit 111 determines the product attribute of the product as a health-related product when the product description of the product includes a predetermined keyword related to health (for example, "glucosamine", "green juice", etc.). The data acquisition unit 111 may acquire the product attribute associated with the product ID included in the purchase data in the storage unit 12. The data acquisition unit 111 specifies the release date, the end date, the normal price, and the product description acquired as described above as sales data.

Further, the data acquisition unit 111 specifies a reference date used for calculating the index (S13). The base date includes a store preferential date (for example, a point-up date or a discount date) and a customer's payday (salary payment date). The data acquisition unit 111 may specify the preferential date based on the acquisition source of the purchase data, or the storage unit 12 may acquire the preferential date associated with the store ID included in the purchase data in advance. The data acquisition unit 111 may specify a preset general payday (for example, the 25th of every month), or the storage unit 12 acquires the payday associated with the customer ID included in the purchase data in advance. May be good.

Next, the index calculation unit 112 calculates an index indicating the customer's purchasing tendency for each customer based on the purchase data acquired by the data acquisition unit 111 from the purchase data storage unit 121. The index calculation unit 112 may calculate one index among the plurality of indexes described below for one customer, or may calculate a plurality of different indexes. That is, in the flowchart of FIG. 5, of steps S14 to S18, the step corresponding to the index to be calculated is performed. Further, the order of steps S14 to S18 may be changed, and at least a part of steps S14 to S18 may be performed in parallel.

First, the index calculation unit 112 calculates the early adopter index, the price sensitive index, the limited product purchase index, the health-related product purchase index, and the premium product purchase index based on the relationship between the sales data and the purchase data. To do.

FIG. 6A is a schematic diagram for explaining the early adopter index. The index calculation unit 112 calculates an early adopter index using the elapsed time from the release date of the product included in the sales data to the purchase date of the product included in the purchase data (S14). Specifically, the index calculation unit 112 is associated with the product ID included in the purchase data acquired by the data acquisition unit 111 (that is, the purchase data of one customer whose index is calculated) from the purchase data storage unit 121. , Acquire purchase data (that is, purchase data of all customers) including a purchase date within a predetermined period (for example, within one year) from the release date of the product.

Next, the index calculation unit 112 arranges the acquired purchase data of all the customers in ascending order of the elapsed time from the release date of the product to the purchase date. The index calculation unit 112 counts the cumulative number of purchase data in ascending order of elapsed time, and early purchase data in which the ratio of the cumulative number to the total number of purchase data of all customers is within a predetermined value (for example, within 16%). Identify as purchase data for the purchase. Such early purchase purchase data corresponds to innovators and early adopters in the innovator theory, and shows a predetermined percentage of purchase data in order of earliest purchase of goods.

Next, the index calculation unit 112 calculates the total purchase amount, which is the sum of the purchase amounts included in the purchase data of one customer acquired by the data acquisition unit 111. In addition, the index calculation unit 112 extracts the purchase data specified as the above-mentioned early purchase from the purchase data. The index calculation unit 112 calculates the ratio obtained by dividing the total value of the purchase amount included in the extracted early purchase purchase data by the calculated total purchase amount. Then, the index calculation unit 112 calculates the value obtained by logarithmically converting the calculated ratio and converting it into an early adopter index as an early adopter index. The index calculation unit 112 may use an early adopter index as a value obtained by performing other processing such as converting the calculated ratio into a deviation value after converting it to a square root. Therefore, by referring to the early adopter index, the user can easily grasp whether or not the customer tends to purchase early when the new product is launched.

FIG. 6B is a schematic diagram for explaining the price-sensitive index. The index calculation unit 112 calculates the price-sensitive index by using the difference between the normal price of the product included in the sales data and the purchase price of the product included in the purchase data (S15). Specifically, the index calculation unit 112 calculates the total purchase amount, which is the sum of the purchase amounts included in the purchase data of one customer acquired by the data acquisition unit 111. In addition, the index calculation unit 112 calculates a discount amount obtained by subtracting the purchase amount included in the purchase data from the normal price of the product included in the sales data for the product with the product ID included in each of the purchase data. The index calculation unit 112 calculates the ratio obtained by dividing the total value of the calculated discount amounts by the calculated total purchase amount.

Then, the index calculation unit 112 calculates a value obtained by logarithmically converting the calculated ratio and converting it into a deviation value as a price-sensitive index. The index calculation unit 112 may use a value that has undergone other processing, such as converting the calculated ratio into a deviation value after performing square root conversion, as a price-sensitive index. Therefore, the user can easily grasp whether or not the customer tends to purchase the discounted product by referring to the price sensitive index.

FIG. 6C is a schematic diagram for explaining a limited product purchase index, a health-related product purchase index, and a premium product purchase index. The index calculation unit 112 extracts a specific attribute product having a specific attribute from the products with the product ID included in the purchase data based on the sales data, and uses the purchase price of the specific attribute product to obtain a limited product purchase index. The health-related product purchase index and the premium product purchase index are calculated (S16).

Specifically, the index calculation unit 112 calculates the total purchase amount, which is the sum of the purchase amounts included in the purchase data of one customer acquired by the data acquisition unit 111. Further, the index calculation unit 112 extracts a product whose product attribute included in the sales data is a limited product from the products having the product ID included in each of the purchase data. The index calculation unit 112 calculates a ratio obtained by dividing the total purchase price of the extracted limited products by the calculated total purchase price. Then, the index calculation unit 112 calculates a value obtained by logarithmically converting the calculated ratio and converting it into a deviation value as a limited product purchase index. The index calculation unit 112 may use a value obtained by performing other processing such as converting the calculated ratio into a deviation value after converting it to a square root as a limited product purchase index. Therefore, the user can easily grasp whether or not the customer tends to purchase a product having a limited sales period or a limited number of sales by referring to the limited product purchase index.

Further, the index calculation unit 112 extracts a product whose product attribute included in the sales data is a health-related product from the products having the product ID included in each of the purchase data. The index calculation unit 112 calculates a ratio obtained by dividing the total purchase price of the extracted health-related products by the calculated total purchase price. Then, the index calculation unit 112 calculates a value obtained by logarithmically converting the calculated ratio and converting it into a deviation value as a health-related product purchase index. The index calculation unit 112 may use a value obtained by performing other processing such as converting the calculated ratio into a deviation value after converting it to a square root as a health-related product purchase index. Therefore, the user can easily grasp whether or not the customer tends to purchase the health-related product by referring to the health-related product purchase index.

Further, the index calculation unit 112 extracts a product whose product attribute included in the sales data is a premium product from the products having the product ID included in each of the purchase data. The index calculation unit 112 calculates the ratio of the total purchase price of the extracted premium products divided by the calculated total purchase price. Then, the index calculation unit 112 calculates the calculated purchase ratio as a premium product purchase index after logarithmically converting the value into a deviation value. The index calculation unit 112 may use a value obtained by performing other processing such as converting the calculated ratio into a deviation value after converting it to a square root as a premium product purchase index. Therefore, the user can easily grasp whether or not the customer tends to purchase a high-class premium product among the same type of products by referring to the premium product purchase index.

Second, the index calculation unit 112 calculates the preferential day purchase index and the payday after-payment purchase index based on the relationship between the base date and the purchase date. FIG. 7A is a schematic diagram for explaining the preferential day purchase index. The index calculation unit 112 calculates the preferential day purchase index based on the relationship between the preferential date of the store indicated by the reference date specified by the data acquisition unit 111 and the purchase date of the product included in the purchase data (S17). ..

Specifically, the index calculation unit 112 calculates the total purchase amount, which is the sum of the purchase amounts included in the purchase data of one customer acquired by the data acquisition unit 111. Further, the index calculation unit 112 extracts the purchase data including the purchase date from the purchase data on the preferential date indicated by the base date. The index calculation unit 112 calculates a ratio obtained by dividing the total value of the purchase amount included in the extracted purchase data on the preferential day by the calculated total purchase amount. Then, the index calculation unit 112 calculates a value obtained by logarithmically converting the calculated ratio and converting it into a deviation value as a preferential day purchase index. The index calculation unit 112 may use a value obtained by performing other processing such as converting the calculated ratio into a deviation value after converting it to a square root as a preferential day purchase index. Therefore, the user can easily grasp whether or not the customer tends to purchase the product on the preferential day such as the point-up date or the discount date by referring to the preferential day purchase index.

FIG. 7B is a schematic diagram for explaining the post-payday purchase index. The index calculation unit 112 calculates the post-payday purchase index based on the relationship between the customer's payday indicated by the reference date specified by the data acquisition unit 111 and the purchase date of the product included in the purchase data (S18). ..

Specifically, the index calculation unit 112 purchases in the purchase data of one customer acquired by the data acquisition unit 111 during a predetermined period (for example, one week before the payday) before the payday indicated by the base date. Extract purchase data that includes days. Further, the index calculation unit 112 includes the purchase date in a predetermined period (for example, one week after the payday) after the payday indicated by the base date in the purchase data of one customer acquired by the data acquisition unit 111. Extract purchasing data. The index calculation unit 112 calculates the ratio of the total value of the purchase amount included in the extracted purchase data after the payday to the total value of the purchase amount included in the extracted purchase data before the payday. Then, the index calculation unit 112 calculates a value obtained by converting the calculated ratio into a square root and converting it into a deviation value as a post-payday purchase index. The index calculation unit 112 may use a value that has undergone other processing, such as logarithmically converting the calculated ratio and then converting it into a deviation value, as a post-payday purchase index. Therefore, the user can easily grasp whether or not the customer tends to purchase more products after the payday by referring to the payday purchase index.

The purchase data and the product data may not include a part of the information shown in FIGS. 3 (a) and 3 (b) depending on the data provider. In such a case, the index calculation unit 112 cannot calculate some of the above-mentioned early adopter index, price sensitive index, limited product purchase index, health-related product purchase index, and premium product purchase index. Therefore, when the index calculation unit 112 cannot calculate a part of the indexes among the plurality of indexes, the index estimation unit 113 estimates the part of the index by using an index other than the part of the index (S19). In other words, for each index contained in the plurality of indexes, when the purchase data and the product data include all the information for calculating the index (that is, when the index calculation unit 112 can calculate the index), the index The calculation unit 112 calculates the index, and the index estimation unit 113 does not estimate the index. On the other hand, for each index included in the plurality of indexes, when the purchase data and the product data do not include at least a part of information for calculating the index (that is, when the index calculation unit 112 cannot calculate the index). , The index calculation unit 112 does not calculate the index, and the index estimation unit 113 estimates the index.

FIG. 8 is a schematic diagram of the index estimation process by the index estimation unit 113. FIG. 8 shows the customer attribute data of the customer and the plurality of indexes calculated by the index calculation unit 112 for the customer in association with each other. In the example of FIG. 8, since the data providers of the customers whose customer IDs are C001 and C002 and the customers whose customer IDs are D001 are different, the index calculation unit 112 has an index γ for the customers whose customer IDs are D001. It represents a state that cannot be calculated.

The index estimation unit 113 generates a model showing the relationship between a plurality of indexes based on the index calculated for the customer whose customer ID is D001 and the customer whose customer ID is similar to the customer attribute data C001. To do. The similarity of the customer attribute data of the customer means that, for example, at least a part of the customer attribute data stored in the customer attribute data storage unit 123 is the same or similar. For example, the model is a mathematical formula that outputs the exponent γ by inputting the exponent α and the exponent β in FIG. The model may be generated using the indices of multiple customers. The model may be pre-generated and stored in the storage unit 12 based on a plurality of indices of the plurality of customers.

The index estimation unit 113 estimates the index γ that the index calculation unit 112 could not calculate by inputting the indexes α and β calculated by the index calculation unit 112 into the generated model for the customer whose customer ID is D001. To do. Alternatively, the index estimation unit 113 calculates the degree of similarity (Euclidean distance, etc.) between the indexes α and β of the customer whose customer ID is C001 and the indexes α and β of the customer whose customer ID is D001. Using the calculated similarity, the index γ of the customer with the customer ID of D001 may be estimated from the index γ of the customer with the customer ID of C001.

As a result, even if a part of the index cannot be calculated by the data provider, the index estimation unit 113 can estimate the index and perform the subsequent processing. The index calculation unit 112 and the index estimation unit 113 store the calculated index and the estimated index in the calculation result storage unit 124 in association with the customer ID.

FIG. 9 is a diagram showing a flowchart of the group index output process S2 executed by the data processing system S according to the present embodiment. The group generation unit 114 generates one or more groups including one or more customers by classifying the customers (S21).

10 (a) and 10 (b) are schematic views of a method in which the group generation unit 114 generates a group. The group generation unit 114 uses the customer attribute data acquired by the data acquisition unit 111 from the customer attribute data storage unit 123, or uses a combination of a plurality of indexes calculated and estimated by the index calculation unit 112 and the index estimation unit 113. , Or both of them to generate a group of customers.

FIG. 10A shows the group G1 generated based on the customer attribute data. The group generation unit 114 generates a group G1 including one or more customers having at least one of the customer's gender, age group, and place of residence common to the customer attribute data stored in the customer attribute data storage unit 123. In the example of FIG. 10A, the group generation unit 114 generates a group G1 of customers having the same gender, age group, and place of residence, but only a part of the gender, age group, and place of residence. May generate a group G1 of customers in common.

In this way, the data processing device 1 generates the customer group G1 based on the customer attribute data, thereby classifying the customers according to the customer attributes such as gender, age, and place of residence, and the purchasing tendency for each customer group. Can output an index indicating.

FIG. 10B represents a group G2 generated based on a combination of a plurality of indices. The group generation unit 114 is one or more persons having a similar combination of two or more predetermined indexes (for example, an early adopter index and a price sensitive index) among a plurality of indexes calculated by the index calculation unit 112 or the index estimation unit 113. Generate group G2 including customers. For example, the group generation unit 114 classifies a customer into one of the group G2 depending on whether or not each of the plurality of indexes of the customer is equal to or higher than a predetermined reference value. In the example of FIG. 10B, the group G2 is generated for each combination of whether or not the index α is equal to or greater than the reference value and whether or not the index β is equal to or greater than the reference value.

The group generation unit 114 is not limited to the method shown here, and may generate the group G2 by performing a known clustering process on a plurality of indices of a plurality of customers. In this way, the data processing device 1 classifies customers by purchasing tendency (personality) by generating a group G2 of customers based on a combination of a plurality of indexes, and further, an index showing a purchasing tendency for each group of customers. Can be output.

The group generation unit 114 assigns group identification information for identifying the group to each of the generated groups G1 and G2. The group specific information is information indicating a combination of customer attribute data or an index used to generate a group. In the example of FIG. 10A, the group-specific information of group G1 is age, gender, and place of residence, and in the example of FIG. 10B, the group-specific information of group G2 is high or low of a plurality of indexes.

The group index calculation unit 115 calculates a statistic regarding the index of one or more customers belonging to each of the one or more groups generated by the group generation unit 114 as the group index of each of the one or more groups (S22). The statistics related to the index are, for example, the average value, the median value, the mode value, and the like of the index calculated by the index calculation unit 112 for one or more customers belonging to the group.

When the index calculation unit 112 calculates a plurality of different indexes, the group index calculation unit 115 calculates the statistics for each of the plurality of indexes as a plurality of different group indexes. That is, the group index calculation unit 115 sets the group index for each group for each of the early adopter index, the price sensitive index, the limited product purchase index, the health-related product purchase index, the premium product purchase index, the preferential day purchase index, and the payday after-sales purchase index. Is calculated. The group index calculation unit 115 stores the calculated group index in the calculation result storage unit 124 in association with the group specific information.

The output unit 116 outputs the calculated group index by the group index calculation unit 115 in association with each of the one or more groups generated by the group generation unit 114 (S23). Specifically, the output unit 116 transmits the display information for displaying the group index and the group in association with each other to the information terminal 2. The information terminal 2 causes the display unit to display an index list screen showing the group index for each group according to the display information received from the data processing device 1.

FIG. 11 is a front view of the information terminal 2 displaying the index list screen. The index list screen includes the group index 21 associated with the group (group specific information). The group index 21 includes one or a plurality of group indexes calculated for each group by the group index calculation unit 115. By referring to the index list screen, the user can grasp the purchasing tendency for each group of customers who have similar customer attribute data such as gender, age, and place of residence, or for each group of customers who have similar combinations of multiple indexes. , Can efficiently analyze customer purchasing behavior.

In the example of FIG. 11, the information terminal 2 displays the information associating the group with the group index, but may display the information associating the customer attribute data of the customer with the group to which the customer belongs. This allows the user to know which group each customer belongs to.

Further, the output unit 116 transmits the display information for displaying the relationship between the group index and the value related to the group different from the group index to the information terminal 2. The display information includes a regression line (correlation coefficient) between the group index and the value related to the group, and the group index and the value related to the group for each group.

The value related to the group is collected separately from the purchase data and stored in the storage unit 12 in advance. For example, the value related to a group is the measure response rate for each group, and the purchase amount, purchase frequency, etc. within the group have improved within a predetermined period after measures such as campaigns have been taken at a specific product or a specific store. The ratio of customers. The information terminal 2 causes the display unit to display an exponential distribution screen showing the relationship between the group index and the value related to the group according to the display information received from the data processing device 1.

FIG. 12 is a front view of the information terminal 2 displaying the exponential distribution screen. The exponential distribution screen includes a two-dimensional plane 22 showing the relationship between the group exponent and the values for the group, and a regression line 23 between the group exponent and the values for the group. Further, the information terminal 2 displays the group 24 (group specific information) at the position of the group index and the value related to the group on the plane 22. By referring to the exponential distribution screen, the user can easily grasp and analyze the relationship between the group index and the values related to the group for each group. For example, referring to the exponential distribution screen of FIG. 12, the user may be able to effectively obtain a response to the measure by focusing on the group of men in their thirties in 1-chome, XXX-cho. Can be judged.

The output unit 116 is not limited to the display on the information terminal 2, and may output the group index by printing on paper using a printer, storing in a storage medium, or transmitting to the outside by communication. ..

[Effect of Embodiment]
The data processing system S according to the present embodiment outputs a group index in which an index indicating a customer's purchasing tendency is statistic for each customer group. Therefore, the user who uses the data processing system S can grasp the purchasing tendency for each group of customers by referring to the group index, and can efficiently analyze the purchasing behavior of the customers.

Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes can be made within the scope of the gist thereof. is there. For example, the specific embodiment of the distribution / integration of the device is not limited to the above embodiment, and all or a part thereof may be functionally or physically distributed / integrated in any unit. Can be done. Also included in the embodiments of the present invention are new embodiments resulting from any combination of the plurality of embodiments. The effect of the new embodiment produced by the combination has the effect of the original embodiment together.

The processor of the data processing device 1 is the main body of each step included in the data processing methods shown in FIGS. 4, 5 and 9. That is, the processor of the data processing device 1 reads a program for executing the data processing methods shown in FIGS. 4, 5 and 9 from the storage unit, executes the program, and controls each part of the data processing device 1. By doing so, the data processing methods shown in FIGS. 4, 5 and 9 are executed. Some of the steps included in the data processing methods shown in FIGS. 4, 5 and 9 may be omitted, the order between the steps may be changed, and a plurality of steps may be performed in parallel.

S Data processing system 1 Data processing device 11 Control unit 112 Index calculation unit 113 Index estimation unit 114 Group generation unit 115 Group index calculation unit 116 Output unit 12 Storage unit 121 Purchase data storage unit 122 Product data storage unit 123 Customer attribute data storage unit

Claims (11)

A data acquisition unit that acquires customer attribute data indicating customer attributes in association with purchase data indicating information regarding the purchase of products by the customer, and a data acquisition unit.
An index calculation unit that calculates an index indicating the purchasing tendency of the customer for each customer based on the purchase data acquired by the data acquisition unit.
When a part of the indexes cannot be calculated for the first customer, the index other than the part of the indexes and the second customer whose customer attribute data are similar to those of the first customer are selected. An estimation unit that estimates a part of the exponents using a model showing the relationship between the plurality of the exponents generated based on the exponents calculated on the index.
A group generation unit that generates one or more groups including one or more customers by classifying the customers using the customer attribute data acquired by the data acquisition unit.
An output unit that outputs statistics on the index of one or more customers belonging to each of the one or more groups as a group index of each of the one or more groups in association with each of the one or more groups.
A data processing device that has. The index calculation unit calculates a plurality of different indexes for each customer based on the purchase data acquired by the data acquisition unit.
The data processing apparatus according to claim 1, wherein the output unit outputs a plurality of the statistics relating to the plurality of the indexes as a plurality of different group indexes. The data processing apparatus according to claim 2, wherein the group generation unit generates the group including one or more customers having a similar combination of a plurality of different indices. Any of claims 1 to 3 , wherein the group generation unit generates the group including one or more customers who share at least one of the customer's gender, age group, and place of residence included in the customer attribute data. The data processing apparatus according to one item. The index calculation unit according to any one of claims 1 to 4 , wherein the index calculation unit calculates the index based on the relationship between the sales data indicating information on the sale of the product by the store and the purchase data. Data processing device. The data processing device according to claim 5 , wherein the index calculation unit calculates the index using the elapsed time from the release date of the product included in the sales data to the purchase date of the product included in the purchase data. .. The index calculation unit calculates the index by using the difference between the normal price of the product included in the sales data and the purchase price of the product included in the purchase data, according to claim 5 or 6 . Data processing equipment. The index calculation unit extracts a specific attribute product having a specific attribute from the product based on the sales data, and calculates the index using the purchase price of the specific attribute product included in the purchase data. , The data processing apparatus according to any one of claims 5 to 7. The index calculation unit calculates the index based on the relationship between the preset reference date and the purchase date of the product included in the purchase data, according to any one of claims 1 to 8. The data processing device described. The output unit corresponds to any one of claims 1 to 9 representing the group on a two-dimensional plane representing the relationship between the group index and a value related to the group different from the group index. The data processing device described. The processor runs,
A step of acquiring the customer attribute data and the purchase data from a storage unit that stores the customer attribute data indicating the customer attributes and the purchase data indicating the information regarding the purchase of the product by the customer in association with each other.
Based on the purchase data acquired by the acquisition step, an index indicating the purchasing tendency of the customer is calculated for each customer, and
When a part of the indexes cannot be calculated for the first customer, the index other than the part of the indexes and the second customer whose customer attribute data are similar to those of the first customer are selected. A step of estimating a part of the exponents using a model showing the relationship between the plurality of the exponents generated based on the exponents calculated on the index.
A step of generating one or more groups including one or more customers by classifying the customers using the customer attribute data acquired by the acquisition step.
A step of outputting statistics on the index of one or more customers belonging to each of the one or more groups as a group index of each of the one or more groups in association with each of the one or more groups.
A data processing method that has.

Images