JP7614140B2 - Information processing device, information processing method, and information processing program - Google Patents

Description

The present invention relates to an information processing device, an information processing method, and an information processing program, and in particular to a technology for predicting potential customers who are expected to use a service.

In order to maximize the effectiveness of advertising, it is important to appropriately extract potential customers (target customers) who are expected to use the advertised service as a group of users to whom the advertisement is delivered. Patent Document 1 discloses a technology that uses machine learning to predict the future purchasing trends of any user. According to this document, it is possible to extract potential customers based on the prediction results.

JP 2021-189983 A

In the above-mentioned document, the future purchasing trends of a given user are predicted using the user's past behavioral data. Specifically, the present is set as a reference timing, and purchasing behavior after this reference timing is predicted. However, the technology in document 1 cannot predict purchasing behavior after an arbitrary reference timing in time, such as a future timing, and cannot predict potential customers after this reference timing.

The present invention has been made in consideration of the above problems, and aims to provide a technology for predicting potential customers after an arbitrary set reference time.

In order to solve the above problem, one aspect of the information processing device according to the present invention has a setting means for setting a reference timing and a target service, an acquisition means for acquiring user characteristics including factual characteristics including user attributes of the user before the reference timing and a service usage history of the target service after the reference timing, and a prediction means for predicting whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing, based on the user characteristics and the service usage history.

In order to solve the above problem, one aspect of the information processing method according to the present invention includes a setting step for setting a reference timing and a target service, an acquisition step for acquiring user characteristics including factual characteristics including user attributes of the user before the reference timing and a usage history of the target service after the reference timing, and a prediction step for predicting whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing, based on the user characteristics and the service usage history.

In order to solve the above problem, one aspect of the program according to the present invention is an information processing program for causing a computer to execute information processing, the information processing program causing the computer to execute processes including a setting process for setting a reference timing and a target service, an acquisition process for acquiring user characteristics including factual characteristics including user attributes of the user before the reference timing and a usage history of the target service after the reference timing, and a prediction process for predicting whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing, based on the user characteristics and service usage history.

According to the present invention, it is possible to predict potential customers after any set reference time.
The above-mentioned objects, aspects, and advantages of the present invention, as well as objects, aspects, and advantages of the present invention not described above, will be understood by those skilled in the art from the following detailed description of the invention by referring to the accompanying drawings and the claims.

FIG. 1 shows an example of the configuration of an information processing system. FIG. 2 shows an example of a functional configuration of the information processing device 10 according to the embodiment. FIG. 3 shows an example of the hardware configuration of the information processing device 10 and the user device 11. As shown in FIG. FIG. 4A shows a conceptual diagram of the learning process of the purchase prediction model 112. FIG. 4B shows a conceptual diagram of a purchase probability prediction process by the purchase prediction model 112. FIG. 5A shows an outline of a flow of a purchase probability prediction process performed by the potential customer prediction unit 105. FIG. 5B shows a schematic flow of a purchase probability prediction process by the potential customer prediction unit 105 when the configuration file 121 is used. FIG. 6 shows a flowchart of the process executed by the information processing device 10.

Below, an embodiment for carrying out the present invention will be described in detail with reference to the attached drawings. Among the components disclosed below, those having the same functions are given the same reference numerals, and their description will be omitted. Note that the embodiment disclosed below is one example of a means for realizing the present invention, and should be appropriately modified or changed depending on the configuration of the device to which the present invention is applied and various conditions, and the present invention is not limited to the following embodiment. Furthermore, not all of the combinations of features described in this embodiment are necessarily essential to the solution of the present invention.

[Example of configuration of information processing system]
Fig. 1 shows an example of the configuration of an information processing system according to this embodiment. As an example, as shown in Fig. 1, this information processing system includes an information processing device 10 and multiple user devices 11-1 to 11-N (N>1) used by multiple arbitrary users 1 to N. In the following description, unless otherwise specified, the user devices 11-1 to 11-N may be collectively referred to as user device 11. In the following description, the terms user device and user may be used synonymously.

The user device 11 is, for example, a device such as a smartphone or a tablet, and is configured to be able to communicate with the information processing device 10 via a public network such as LTE (Long Term Evolution) or a wireless communication network such as a wireless LAN (Local Area Network). The user device 11 has a display unit (display surface) such as a liquid crystal display, and each user can perform various operations using a GUI (Graphic User Interface) equipped on the liquid crystal display. The operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or a stylus.
The user device 11 is not limited to the device shown in Fig. 1, but may be a device such as a tablet terminal or a notebook PC. The user device 11 may also be provided with a separate display screen.

The user device 11 can use web services (Internet-related services) provided from the information processing device 10 or from another device (not shown) via the information processing device 10. The web services can include online malls, online supermarkets, or communication, finance, real estate, sports, and travel reservation services provided via the Internet. In order to use the web services, the user device 11 can register information such as the user's address, the user's name, the credit card number held by the user, and the user's demographic information (demographic user attributes such as gender, age, residential area, occupation, and family structure).

[Functional configuration of information processing device 10]
The information processing device 10 according to the present embodiment sets a target service (target service) for predicting potential customers. The information processing device 10 also sets a reference timing (hereinafter, reference timing) for predicting potential customers, a certain period (first period) before (past) the reference timing, and a certain period (second period) after (future) the reference timing. The information processing device 10 predicts the purchase probability of the target user for the target service in the second period using user data in the first period acquired from any user (target user) among multiple users 1 to N. The information processing device 10 further predicts whether the target user is a potential customer or not based on the predicted purchase probability. In this embodiment, a potential customer is a user who is expected to use the target service, and the target service is any web service or any item available to the user, and can be read as a target item. In this embodiment, "use" means "purchase", but may include any form of use such as registration and membership.

In this embodiment, the items available to the user may be tangible or intangible things (Things) that can be provided for various services. For example, for financial (fintech) services, there are items such as bank accounts, financial products such as stocks, investment trusts, and insurance products, cryptocurrencies, and smartphone app payments. For digital content services, there are items such as video content such as movies and animations, and still image content such as photos, illustrations, and text. For e-commerce services, there are items such as intangible or tangible products handled in online shopping. For travel services, there are items such as information and reservations regarding hotels, package tours, and transportation. For mobile services, there are items such as mobile devices, public network/Internet connections, and communication usage fees. For advertising and media services, there are items such as offline and offline advertising products, direct mail, and advertising via broadcasting and the Internet. For card services, there are items such as credit card payments and point transactions. For sports and cultural services, there are items such as events such as sporting events and concerts, and products sold at events.

FIG. 2 is a block diagram showing an example of a functional configuration of the information processing device 10 according to the present embodiment.
2 includes a user data acquisition unit 101, a time information setting unit 102, a service setting unit 103, a user data selection unit 104, a potential customer prediction unit 105, an output unit 106, a learning model storage unit 110, and a file storage unit 120. The learning model storage unit 110 is configured to be capable of storing a user feature prediction model 111 and a purchase prediction model 112. In addition, the file storage unit 120 is configured to be capable of storing a configuration file 121.

The user data acquisition unit 101 acquires data (user data) indicating characteristics (user characteristics) of the user device and the user from each of the user devices 11-1 to 11-N. In this embodiment, the user data includes the user's actual characteristics and the user's estimated characteristics that are estimated based on the user's actual characteristics.

The factual characteristics of a user are characteristics obtained from factual information about the user device and the user. For example, the user data acquisition unit 101 can acquire the factual characteristics of the user directly from the user device 11. Furthermore, the user data acquisition unit 101 can acquire the factual characteristics of the user as information registered in a web service by the user of the user device 11. The factual characteristics of a user include factual attributes about the user device and the user (hereinafter, user attributes). Furthermore, the factual characteristics of a user may include tracking data (tracking log) and usage history data.

User attributes are attributes related to the user himself/herself, including the IP address of the user device, the user's address and name, the credit card number held by the user, and the user's demographic information (demographic attributes such as gender, age, residential area, occupation, and family structure). User attributes may also include the registration number and registered name when using a specific web service.

Tracking data is data (data indicating behavior on a website) obtained by tracking a user's behavior on a website that provides a service (e.g., multiple services including the target service). The tracking data includes, for example, data on behavior on the website, such as the number of clicks, the route taken to visit the site, the number of clicks on items on the site, or the time spent viewing the site.

The usage history data includes service usage history (usage status, search history, item purchase history (including purchase results)) when the user uses a service (e.g., multiple services including the target service). The usage history data may also include data such as the amount spent on the service used, the number of items purchased, or the number of items registered. The usage history data may also include information regarding points that can be accumulated by using the web service.

The estimated features of a user are estimated (derived) from the user's factual features. In this embodiment, the user data acquisition unit 101 can apply the factual features of the user to the trained user feature prediction model 111 to acquire the estimated features. The user feature prediction model 111 is configured to estimate and output the probability (correspondence probability) that each of a plurality of user features corresponds (matches) the target user, using the factual features of the target user as input. The user data acquisition unit 101 determines the estimated features of the target user based on the correlation probability.

For example, the user data acquisition unit 101 inputs the factual characteristics of the target user into the user characteristic prediction model 111. The user characteristic prediction model 111 outputs the probability for each of a number of items that the target user is estimated to purchase and a number of preferences that the target user may have as the relevant probability. The user data acquisition unit 101 can then acquire, as the estimated characteristics of the target user, estimated user attributes that correspond to items and preferences that have a probability equal to or greater than a predetermined value.

Note that the estimated features of a user are not limited to features estimated beyond the present (i.e., the future). A standard for feature estimation may be set before the present (i.e., the past), and the estimated features of the user after the standard may be estimated.

The time information setting unit 102 sets, as time information, a reference timing for predicting potential customers, a certain period (first period) after (past) the reference timing, and a certain period (second period) before (future) the reference timing. The time information may be set by an operator through an input operation using an input unit (input unit 35 in FIG. 3), may be set in advance in the system, or may be set by any program stored in a storage unit (ROM 32 or RAM 33 in FIG. 3).

The service setting unit 103 sets the service (target service) that is the target of the potential customer prediction. The target service may be set by an operator through an input operation using the input unit (input unit 35 in FIG. 3), may be set in advance in the system, or may be set by any program stored in the storage unit (ROM 32 or RAM 33 in FIG. 3).

The user data selection unit 104 is an optional function that selects user data to be used in the prediction process by the potential customer prediction unit 105 from the user data acquired by the user data acquisition unit 101. In this embodiment, the user data selection unit 104 can select user data based on a configuration file 121 stored in the file storage unit 120. The configuration file 121 will be described later.

The prospective customer prediction unit 105 predicts and extracts prospective customers using the trained purchase prediction model 112. Specifically, the prospective customer prediction unit 105 predicts the purchase probability of the target user for the target service in a second time period using user data for a first time period acquired from any user (target user) among the multiple users 1 to N. Then, the prospective customer prediction unit 105 predicts whether the target user is a prospective customer or not based on the purchase probability. Furthermore, the prospective customer prediction unit 105 can generate a group of prospective customers (user segment) among users 1 to N by predicting whether each of users 1 to N is a prospective customer or not.

The output unit 106 outputs information on the potential customers predicted by the potential customer prediction unit 105. The output unit 106 may generate and output information on the potential customers. The output unit 106 may also generate and output (provide) advertisements related to the target service to user segments. The output may be any output process, and may be output to an external device via a communication I/F (communication I/F 37 in FIG. 3) or may be displayed on a display unit (display unit 36 in FIG. 3).

The learning model storage unit 110 stores the learned user feature prediction model 111 described above. The learning model storage unit 110 also stores a purchase prediction model 112. The purchase prediction model 112 will be described later.

The file storage unit 120 stores a configuration file 121. The configuration file 121 is set with setting information for each of a plurality of services for predicting potential customers. In this embodiment, the configuration file 121 includes information indicating data necessary for predicting potential customers for each service, among the user data that can be acquired by the user data acquisition unit 101. The user data selection unit 104 can select user data necessary for predicting potential customers for the target service from the user data acquired by the user data acquisition unit 101 according to the configuration file 121. As a result, only the user data necessary for predicting potential customers for the target service can be selected from the huge amount of data acquired by the user data acquisition unit 101, so that the data selection process by the potential customer prediction unit 105 is omitted and the processing load is reduced.

The configuration file 121 according to this embodiment also includes settings for the purchase prediction model 112 for each service. For example, the configuration file 121 includes parameters to be used for the target service, a model to be used (such as CatBoost, as described below), the structure of the user data to be used, and the like. This allows the potential customer prediction unit 105 to perform potential customer prediction processing using a purchase prediction model 112 that is more suited to the target service.

[Hardware configuration of information processing device 10]
FIG. 3 is a block diagram showing an example of a hardware configuration of the information processing device 10 according to the present embodiment.
The information processing apparatus 10 according to the present embodiment can be implemented on any single or multiple computers, mobile devices, or any other processing platform.
3, the information processing device 10 is illustrated as being implemented in a single computer, but the information processing device 10 according to the present embodiment may be implemented in a computer system including multiple computers. The multiple computers may be connected to each other via a wired or wireless network so as to be able to communicate with each other.

3, the information processing device 10 may include a CPU 31, a ROM 32, a RAM 33, a HDD 34, an input unit 35, a display unit 36, a communication I/F 37, and a system bus 38. The information processing device 10 may also include an external memory.
A CPU (Central Processing Unit) 31 generally controls the operations of the information processing device 10, and controls each of the components (32 to 37) via a system bus 38, which is a data transmission path.

The ROM (Read Only Memory) 32 is a non-volatile memory that stores control programs and the like necessary for the CPU 31 to execute processing. Note that the programs may be stored in a non-volatile memory such as a HDD (Hard Disk Drive) 34 or an SSD (Solid State Drive) or in an external memory such as a removable storage medium (not shown).
The RAM (Random Access Memory) 33 is a volatile memory and functions as a main memory, a work area, etc. of the CPU 31. That is, when executing a process, the CPU 31 loads necessary programs, etc. from the ROM 32 into the RAM 33 and executes the programs, etc. to realize various functional operations. In this embodiment, the learning model storage unit 110 and the file storage unit 120 can be configured by the RAM 33.

The HDD 34 stores, for example, various data and various information required when the CPU 31 performs processing using a program. The HDD 34 also stores, for example, various data and various information obtained when the CPU 31 performs processing using a program.
The input unit 35 is composed of a keyboard and a pointing device such as a mouse.
The display unit 36 is configured with a monitor such as a liquid crystal display (LCD). The display unit 46 may be configured in combination with the input unit 35 to function as a GUI (Graphical User Interface).

The communication I/F 37 is an interface that controls communication between the information processing device 10 and external devices.
The communication I/F 37 provides an interface with a network and executes communication with an external device via the network. Various data, various parameters, etc. are transmitted and received between the external device and the communication I/F 37. In this embodiment, the communication I/F 37 may execute communication via a wired LAN (Local Area Network) or a dedicated line that complies with a communication standard such as Ethernet (registered trademark). However, the network that can be used in this embodiment is not limited to this, and may be configured as a wireless network. This wireless network includes wireless PANs (Personal Area Networks) such as Bluetooth (registered trademark), ZigBee (registered trademark), and UWB (Ultra Wide Band). It also includes wireless LANs (Local Area Networks) such as Wi-Fi (Wireless Fidelity) (registered trademark) and wireless MANs (Metropolitan Area Networks) such as WiMAX (registered trademark). It also includes wireless WANs (Wide Area Networks) such as LTE/3G, 4G, and 5G. Note that the network only needs to be able to connect devices to each other and communicate with each other, and the communication standard, scale, and configuration are not limited to those described above.

At least some of the functions of each element of the information processing device 10 shown in FIG. 2 can be realized by the CPU 31 executing a program. However, at least some of the functions of each element of the information processing device 10 shown in FIG. 2 may be operated as dedicated hardware. In this case, the dedicated hardware operates under the control of the CPU 31.

[Hardware Configuration of User Device 11]
The hardware configuration of the user device 11 shown in Fig. 1 may be the same as that of Fig. 3. That is, the user device 11 may include a CPU 31, a ROM 32, a RAM 33, a HDD 34, an input unit 35, a display unit 36, a communication I/F 37, and a system bus 38. The user device 11 can display various information provided by the information processing device 10 on the display unit 36 and perform processing corresponding to an input operation received from a user via a GUI (configured by the input unit 35 and the display unit 36).

[Prospective customer prediction procedure]
In this embodiment, the potential customer prediction unit 105 applies the user data of the target user acquired by the user data acquisition unit 101 to the trained purchase prediction model 112 to predict a purchase probability. Then, the potential customer prediction unit 105 predicts whether the target user is a potential customer for the target service from the purchase probability. Regarding the procedure for predicting potential customers, first, the purchase prediction model 112 will be described, and then the process of predicting a purchase probability using the purchase prediction model 112 and the process of predicting (extracting) potential customers will be described.

The purchase prediction model 112 is a learning model for machine learning (machine learning model), for example, a learning model based on CatBoost. Alternatively, the purchase prediction model 112 may be another boosting-based learning model such as XGBoost or LightGBM.

The purchase prediction model 112 has been trained using training data based on past user data (user data obtained in the past) acquired by the user data acquisition unit 101. The training process of the purchase prediction model 112 will be described with reference to FIG. 4A. FIG. 4A shows a conceptual diagram of the training process of the purchase prediction model 112. A first period 41 before (past) the reference timing T1 and a second period 42 after (future) the reference timing T1 are set. For example, the first period 41 is one year, and the second period 42 is one month. As the training data, training data obtained by labeling a purchase result (e.g., 0: did not purchase, 1: purchased) corresponding to a service usage history (usage history, purchase history) obtained from usage history data (part of the past user data) for a user feature corresponding to past user data (user data acquired in the first period 41) for an arbitrary service in the first period 41 is prepared. Then, the purchase prediction model 112 is trained using the training data. The reference timing T1, the first period 41, and the second period 42 are arbitrarily set during the learning stage. The reference timing T1 set during purchase prediction (reference timing T1 in the learning data) may further include, for example, information indicating events such as seasons and holidays in addition to time-related information such as date.

The purchasing result reflected in the learning data by labeling refers to the purchasing result that can be confirmed from the usage history data at least in the second period 42. The purchasing result reflected in the learning data by labeling may also refer to the purchasing result that can be confirmed from the usage history data in a continuous period including at least the reference timing T1. The purchasing result reflected in the learning data by labeling may also refer to the difference (time-series change in the purchasing situation) between the purchasing result that can be confirmed from the usage history data in the first period 41 and the purchasing result that can be confirmed from the usage history data in the second period 42. Specifically, the purchasing result reflected in the learning data by labeling may refer to the presence or absence of a positive time-series change in the purchasing situation based on the reference timing T1 based on the presence or absence of a purchase (purchase) in the first period 41 and the presence or absence of a purchase in the second period 42. Here, in the case where there is no purchase (purchase) in the first period 41 and there is a purchase (purchase) in the second period 42, the learning data is positively labeled, for example. Also, here, the learning data is negatively labeled, for example, if there is no purchase in the first period 41 and there is a purchase in the second period 42.

The learning data of the purchase prediction model 112 may further be labeled with the presence or absence of a time series change (temporal change) of the user features in the first period 41 with reference to a reference timing T1 from the first period 41 to the second period 42. For example, the learning data may be labeled with the presence of a time series change of the user features in the first period 41, taking into consideration that when the reference timing T1 is the user's birthday, the age change of the user occurs with reference to the reference timing T1. The learning data may be labeled with the presence or absence of a time series change of the user features for each attribute.

The trained purchase prediction model 112 is configured to input user data for a first period prior to an arbitrary reference timing, and output a purchase probability (purchase score) for a set target service for a second period after the reference timing. The purchase probability is expressed as a value between 0 and 1, and in this embodiment, a purchase probability of 1 indicates that there is a 100% possibility that the user will make a purchase in the target service, and a purchase probability of 0 indicates that there is a 0% possibility that the user will make a purchase in the target service. The trained purchase prediction model 112 is stored in the learning model storage unit 110.

Next, the process of predicting the purchase probability 55 using the purchase prediction model 112 will be described with reference to Figures 4B and 5A. Figure 4B shows a conceptual diagram of the purchase probability prediction process using the purchase prediction model 112, and Figure 5A shows a general flow of the purchase probability prediction process using the potential customer prediction unit 105.

First, the potential customer prediction unit 105 acquires the time information 52 set by the time information setting unit 102 and the target service 53 for potential customer prediction set by the service setting unit 103. The time information 52 includes information on the reference timing T2, the first period 43, and the second period 44 shown in FIG. 4B. In addition, the potential customer prediction unit 105 acquires user data 51 for the first period 43 from the user data acquired by the user data acquisition unit 101.

Next, the potential customer prediction unit 105 inputs the user data 51 into the purchase prediction model 112 and predicts the purchase probability 55 for the target service 53 in the second period 44. Here, the purchase prediction model 112 used is a model trained based on the reference timing T2, the first period 43, and the second period 44. Alternatively, the purchase prediction model 112 used is a model selected based on the reference timing T2, the first period 43, and the second period 44. If the reference timing T2 is set before the present (in the past), the purchase prediction model 112 already trained based on the reference timing T2, the first period 43, and the second period 44 may be used. On the other hand, if the reference timing T2 is set after the present (in the future), there is substantially no training data (i.e., purchase results). In this case, for example, the purchase prediction model 112 already trained based on the reference timing T2, the first period 43, and the second period 44 corresponding to the same date and time one year ago may be selected and used.

The potential customer prediction unit 105 may predict multiple purchase probabilities 55 of the target user based on the purchase prediction models 112 corresponding to each of the multiple combinations of the reference timing T2, the first period 43, and the second period 44, and may determine the final purchase probability 55 based on the multiple purchase probabilities 55 predicted in each purchase prediction model 112. Here, the potential customer prediction unit 105 predicts multiple purchase probabilities 55 after imposing certain constraints on the combinations of the reference timing, the first period, and the second period. The constraints are based on a specified timing, such as the timing of providing an advertisement, and the reference timing, the first period, and the second period. Also, as an example of the constraints, a condition may be a combination in which the second period includes a timing that is the same as the timing of providing an advertisement.

In this way, by inputting user data up to the reference timing T2, specifically, user characteristics in the first period 43, into the purchase prediction model 112, it is possible to predict the user's purchase probability 55 in the second period 44.

By performing this type of purchase probability prediction process for users 1 to N, the potential customer prediction unit 105 can predict the purchase probability 55 of users 1 to N for the target service 53 during the second period 44.

When the configuration file 121 is used, the purchase probability prediction process can be performed more efficiently. FIG. 5B shows an outline flow of the purchase probability prediction process by the potential customer prediction unit 105 when the configuration file 121 is used. As described above, the configuration file 121 includes information indicating data required to predict potential customers for each service, and settings of the purchase prediction model 112 for each service. The potential customer prediction unit 105 acquires user data 51 for the first period 43, and selects data required to predict potential customers for the target service 53 from the user data 51 according to the configuration file 121 (selected user data 54). The potential customer prediction unit 105 also sets the purchase prediction model 112 for the target service 53 according to the configuration file 121.

If the target service 53 is, for example, a travel reservation service, the user data 54 selected according to the configuration file 121 includes the number of travel packages and hotel reservations, the travel destinations, and the costs of the trip in the first period 43. In this way, the user data 51 is a huge amount of data, but by using the configuration file 121, the user data 54 selected corresponding to the target service 53 can be input into the purchase prediction model 112.

When the purchase probability 55 is predicted, the potential customer prediction unit 105 predicts potential customers from the purchase probability 55. Specifically, the potential customer prediction unit 105 can predict one or more users whose purchase probability 55 is equal to or greater than a predetermined threshold as potential customers (users who are expected to use the target service 53 (have a high probability of using it)). The threshold is, for example, 0.5.

Furthermore, the potential customer prediction unit 105 generates a user group consisting of one or more users predicted as potential customers as a user segment. Users included in the user segment have a high probability of using the target service 53. Therefore, the potential customer prediction unit 105 can determine the user segment as a user group suitable as a target for advertisement delivery of the target service 53.

[Process flow]
Fig. 6 shows a flowchart of the process executed by the information processing device 10 according to this embodiment. The process shown in Fig. 6 can be realized by the CPU 31 of the information processing device 10 loading a program stored in the ROM 32 or the like into the RAM 33 and executing it. For the explanation of Fig. 6, the information processing system shown in Fig. 1 will be referred to. It is assumed that the trained user feature prediction model 111 and purchase prediction model 112 are stored in the learning model storage unit 110. Furthermore, the order of the process in Fig. 6 is not limited to the order shown in the figure.

In S61, the time information setting unit 102 sets a reference timing for predicting potential customers, a first period after the reference timing (past), and a certain period before the reference timing (future) (second period). The reference timing is not limited to the present timing, and can be set to any timing in the past or future.

In S62, the service setting unit 103 sets the service (target service 53) that is the target of the potential customer prediction. In this embodiment, the service is a web service or an item that can be used by the user.

In S63, the user data acquisition unit 101 acquires user data from users (target users) who are the targets of the potential customer prediction. In this embodiment, as described above, the user data (user characteristics) includes the factual characteristics of the user and the estimated characteristics of the user estimated based on the factual characteristics of the user.

In S64, the potential customer prediction unit 105 inputs the user data acquired in S63 into the purchase prediction model 112 to predict the purchase probability 55 of the target user for the target service 53 set in S62. The purchase probability prediction process is as described above with reference to Figures 4B, 5A, and 5B.

In S65, the potential customer prediction unit 105 determines whether the target user is a potential customer (a user who is expected to use the target service 53) based on the purchase probability 55 of the target user predicted in S64. For example, if the purchase probability 55 is equal to or greater than a predetermined threshold, the potential customer prediction unit 105 determines that the target user is a potential customer.

The information processing device 10 performs the processes of S63 to S65 for users 1 to N. That is, the information processing device 10 performs the processes of S63 to S65 for each of users 1 to N as a target user. Next, in S66, the potential customer prediction unit 105 generates a user group consisting of one or more users of users 1 to N who have been predicted as potential customers, as a user segment.

S67 is an optional process, in which the output unit 106 provides an advertisement for the target service 53 set in S62 to the user segment generated in S66. For example, if the reference timing set in S61 is one month from now, the output unit 106 provides an advertisement for the target service 53 to the user segment for a certain period of time before the reference timing. The provision of the advertisement can have an effective effect on the purchasing behavior of the user segment after the reference timing.

In this way, according to this embodiment, by taking into account the time series changes in user characteristics based on an arbitrarily set reference timing, it is possible to predict with greater accuracy whether a user is a potential customer. In particular, user characteristics are likely to change after life events (marriage, childbirth, etc.), but in this embodiment, by taking into account the time series changes in user characteristics for a certain period of time prior to the reference timing, it is possible to effectively predict whether a user will be a potential customer (in the future).

In this embodiment, an example is shown in which "use" corresponds to "purchase," and the purchase probability 55 is predicted using the purchase prediction model 112, but this embodiment can be applied to any form of use. In other words, a usage prediction model corresponding to any form of use may be used, and potential customers may be predicted according to the usage probability predicted by the usage prediction model.

Below, we will explain several use cases according to the reference timing settings. For this explanation, we will refer to the information processing system shown in Figure 1, the conceptual diagram of the purchase probability prediction process shown in Figure 4B, and the general flow of the purchase probability prediction process shown in Figure 5A.

[First Use Case]
The first use case is a case where the reference timing T2 is set before the present. In this use case, the reference timing T2 is set to the date of a past customer attraction event, and the target service 53 is set to a service related to the area where the customer attraction event was held. The potential customer prediction unit 105 can predict, using the learned purchase prediction model 112, potential customers for the service related to the area among the multiple users 1 to N in the second period 44 after the reference timing T2. The predicted information on the potential customers can be used when a customer attraction event is held again in the area. For example, when a customer attraction event is held in the area in the future, it becomes possible to effectively plan the fulfillment of the service related to the area.

[Second Use Case]
The second use case is a case where a plurality of reference times are set from the past to the future. In this use case, the reference time T2 is set to the convergence timing of a plurality of infectious disease epidemics, and the target service 53 is set to an Internet mall. The potential customer prediction unit 105 can predict potential customers for the Internet mall in a second period 44 after each reference time T2 among the plurality of users 1 to N using the learned purchase prediction model 112. Information on the predicted potential customers can be used for a listing plan according to the convergence timing of the infectious disease epidemic. In other words, the purchasing tendency of users according to different convergence timings is quantified, and it becomes possible to effectively carry out listing plans for the Internet mall.

[Third Use Case]
The third use case is a case where the reference timing T2 is set to a time later than the present. In this use case, the reference timing T2 is set to December 1 of this year (after the reference timing T2), and the target service 53 is set to a restaurant reservation service. The potential customer prediction unit 105 can use the learned purchase prediction model 112 to predict which of the multiple users 1 to N will be potential customers for the restaurant reservation service in the second period 44 after the reference timing T2 and including Christmas. In other words, it can predict whether the multiple users 1 to N will be potential customers for the restaurant reservation service in the future.

Although specific embodiments have been described above, these are merely examples and are not intended to limit the scope of the present invention. The devices and methods described herein may be embodied in forms other than those described above. Furthermore, the above-described embodiments may be omitted, substituted, or modified as appropriate without departing from the scope of the present invention. Forms in which such omissions, substitutions, or modifications have been made are included within the scope of the claims and their equivalents, and belong to the technical scope of the present invention.

The disclosure of this embodiment includes the following configuration.
[1] An information processing device comprising: a setting means for setting a reference timing and a target service; an acquisition means for acquiring user characteristics including factual characteristics including user attributes of a user before the reference timing, and a service usage history of the target service after the reference timing; and a prediction means for predicting, based on the user characteristics and service usage history, whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing.

[2] The information processing device described in [1], characterized in that the user features include estimated features including the user attributes estimated based on the fact features.

[3] The information processing device described in [2], characterized in that the user characteristics further include data indicating the user's actions on a website for providing multiple services including the target service as fact characteristics.

[4] The information processing device according to any one of [1] to [3], characterized in that the prediction means predicts whether the target user is the potential customer by inputting the user characteristics and service usage history into a trained machine learning model.

[5] The information processing device described in [4] further comprises a generation means for generating a group of the potential customers, the acquisition means acquires the user characteristics for each of the multiple target users, the prediction means predicts whether each of the multiple target users is the potential customer, and the generation means generates, as the group of the potential customers, a group consisting of one or more of the multiple target users predicted by the prediction means to be the potential customer.

[6] The information processing device described in [5], further comprising a provision means for providing advertisements related to the target service to the group of potential customers.

[7] An information processing method comprising: a setting step for setting a reference timing and a target service; an acquisition step for acquiring user characteristics including factual characteristics including user attributes of a user before the reference timing and a service usage history of the target service after the reference timing; and a prediction step for predicting whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing, based on the user characteristics and the service usage history.

[8] An information processing program for causing a computer to execute information processing, the information processing program causing the computer to execute processes including a setting process for setting a reference timing and a target service, an acquisition process for acquiring user characteristics including factual characteristics including user attributes of a user before the reference timing and a service usage history of the target service after the reference timing, and a prediction process for predicting whether or not an arbitrary target user different from the user is a potential customer who is expected to use the target service after the reference timing, based on the user characteristics and the service usage history.

1 to N: User, 10: Information processing device, 11-1 to 11-N: User device, 101: User data acquisition unit, 102: Time information setting unit, 103: Service setting unit, 104: User data selection unit, 105: Prospective customer prediction unit, 106: Output unit, 110: Learning model storage unit, 111: User feature prediction model, 112: Purchase prediction model, 120: File storage unit, 121: Configuration file

Claims (7)

A setting means for setting a target reference timing, which is any timing in any year , and a target service;
An acquisition means for acquiring user features including factual features including user attributes of the target user in the past from the target reference timing;
a prediction means for predicting whether the target user is a prospective customer who is expected to use the target service after the target reference time;
having
The prediction means includes:
If the target reference time is in the past, predicting whether the target user is the potential customer by inputting user features of the target user in the past than the target reference time into a learning model trained using user features including fact features including user attributes of an arbitrary user in the past than the target reference time and a service history of the target service by the arbitrary user in the future than the target reference time;
When the target reference timing is in the future, an information processing device predicts whether or not the target user is a potential customer by inputting user characteristics of the target user in the past before the target reference timing into a learning model trained using user characteristics of the arbitrary user in the past before the target reference timing in a year prior to the present, and a service history of the target service by the arbitrary user in the future before the target reference timing in the past year . The information processing apparatus according to claim 1 , wherein the user features of the target user include estimated features including user attributes estimated based on factual features of the target user . The information processing device according to claim 2, characterized in that the user characteristics of the target user further include data indicating the target user's behavior on a website for providing multiple services including the target service as a fact characteristic. A generating means for generating the group of potential customers,
The acquisition means acquires the user characteristics for each of the plurality of target users,
The prediction means predicts whether each of the plurality of target users is a potential customer or not;
The information processing device according to claim 1, characterized in that the generation means generates, as the group of potential customers, a group consisting of one or more of the target users who are predicted to be potential customers by the prediction means. 5. The information processing apparatus according to claim 4 , further comprising providing means for providing an advertisement relating to the target service to the group of potential customers. An information processing method executed by an information processing device,
A setting step of setting a target reference timing, which is an arbitrary timing in an arbitrary year , and a target service;
An acquisition step of acquiring user features including factual features including user attributes of the target user in the past from the target reference timing;
a prediction step of predicting whether the target user is a prospective customer who is expected to use the target service after the target reference time;
having
In the prediction step,
If the target reference time is in the past, predicting whether the target user is the potential customer by inputting user features of the target user in the past than the target reference time into a learning model trained using user features including fact features including user attributes of an arbitrary user in the past than the target reference time and a service history of the target service by the arbitrary user in the future than the target reference time;
When the target reference timing is in the future, an information processing method is provided, characterized in that it predicts whether or not the target user is a potential customer by inputting user characteristics of the target user in the past before the target reference timing into a learning model trained using user characteristics of the arbitrary user in the past year before the target reference timing and a service history of the target service by the arbitrary user in the future before the target reference timing in the past year . An information processing program for causing a computer to execute information processing, the information processing program comprising:
A setting process for setting a target reference timing, which is an arbitrary timing in an arbitrary year , and a target service;
An acquisition process of acquiring user features including fact features including user attributes of the target user in the past from the target reference timing;
a prediction process for predicting whether the target user is a potential customer who is expected to use the target service after the target reference time,
The prediction process includes:
If the target reference time is in the past, predicting whether the target user is the potential customer by inputting user features of the target user in the past than the target reference time into a learning model trained using user features including fact features including user attributes of an arbitrary user in the past than the target reference time and a service history of the target service by the arbitrary user in the future than the target reference time;
When the target reference timing is in the future, predicting whether the target user is the potential customer by inputting user characteristics of the target user in the past before the target reference timing into a learning model trained using user characteristics of the arbitrary user in the past year before the target reference timing and a service history of the target service by the arbitrary user in the past year before the target reference timing.
Information processing program.

Images