JP7633440B2 - Information processing device, information processing method, and information processing program - Google Patents

Description

This disclosure relates to an information processing device, an information processing method, and an information processing program.

For example, Patent Document 1 describes a simulation device that predicts the effectiveness of sending direct mail for a sales promotion campaign for a specific brand of product.

This simulation device has a planned budget receiving section, a customer master table, a condition receiving section, a number of recipients calculation section, a number of recipients presentation section, a number of recipients planned to be sent to receiving section, an expected conversion rate receiving section, an expected number of recipients calculation section, a unit price table, an expected earned value calculation section, an expected ROI calculation section, and a forecast presentation section.

The planned budget receiving unit receives input of the planned budget for the campaign. The customer master table includes attributes and a transfer probability for each customer. The condition receiving unit receives settings of direct mail delivery conditions. The possible number of recipients calculation unit extracts customers who match the delivery conditions from the customer master table and calculates the possible number of recipients for the direct mail. The possible number of recipients presentation unit presents the possible number of recipients. The planned number of recipients reception unit receives settings of the planned number of recipients to whom the direct mail is to be sent, with the possible number of recipients as the upper limit. The expected transfer rate receiving unit receives input of the expected transfer rate according to the delivery conditions. The expected transfer number calculation unit calculates the expected number of transferees from the planned number of recipients and the expected transfer rate. The unit price table stores the acquired value unit price per person, which is set for each attribute of the product brand and indicates the value acquired due to the change in the brand that the customer mainly purchases before and after the campaign. The expected acquired value calculation unit refers to the unit price table, obtains the acquired value unit price by transferring to a specific brand corresponding to the attributes of the product brand that matches the delivery conditions, and calculates the expected acquired value from the expected number of transferees. The expected ROI calculation unit calculates the expected ROI from the planned budget and the expected acquisition value. The prediction presentation unit presents the expected acquisition value and the expected ROI.

Traditionally, sales strategies and other marketing have mainly been conducted using direct mail such as that described in Patent Document 1, as well as stickers and postcards. However, in recent years, with the spread of smartphones and cashless payments, attention has been focused on "digital marketing (sales promotion using digital technology)," which allows people to easily enter campaigns using communication devices and receive prizes and points.

International Publication No. 2007/037178

However, according to the inventor's intensive research, there is room for improvement in conventional digital marketing in terms of predicting and verifying its effectiveness, as well as its utilization.

Therefore, one of the objectives of the present disclosure is to provide an information processing device, an information processing method, and an information processing program that can effectively predict, verify, and utilize the effects of digital marketing.

An information processing device according to one embodiment of the present disclosure has an identification information acquisition unit that acquires identification information of a potential participant in a project, an attribute information acquisition unit that acquires attribute information of the potential participant, a behavior history information acquisition unit that acquires behavior history information of the potential participant, a detailed information acquisition unit that acquires detailed information of the project, and a calculation mode control unit that controls a calculation mode of an effect of the potential participant participating in the project, wherein the calculation mode control unit calculates the effect of the potential participant participating in the project based on a first effect obtained by fitting the identification information, the attribute information, the behavior history information, and the detailed information to a machine learning model, and a second effect obtained by fitting the detailed information to a linear regression model .
An information processing device according to one embodiment of the present disclosure has an identification information acquisition unit that acquires identification information of a candidate participant in a project, an attribute information acquisition unit that acquires attribute information of the candidate participant, a behavior history information acquisition unit that acquires behavior history information of the candidate participant, a detailed information acquisition unit that acquires detailed information about the project, and a calculation mode control unit that controls a calculation mode of an effect of the candidate participant participating in the project, wherein the calculation mode control unit calculates, as the effect of the candidate participant participating in the project, the probability that the candidate participant will participate in the project and whether or not the probability exceeds a predetermined threshold, as well as an output value of a linear regression relational equation, and corrects the predetermined threshold and parameters of the linear regression relational equation based on panel data conducted by a survey company.

According to one aspect of the present disclosure, it is possible to provide an information processing device, an information processing method, and an information processing program that can effectively predict, verify, and utilize the effects of digital marketing.

FIG. 1 is a diagram illustrating an example of a schematic configuration of an information processing device. 2 is a block diagram showing an example of an internal configuration of an information processing server; FIG. 1 is a conceptual diagram showing an example of identification information, attribute information, behavior history information, detailed information, and a purchase probability based thereon. FIG. 13 is a diagram illustrating an example of effect calculation based on a machine learning model. FIG. 13 is a diagram illustrating an example of effect calculation based on a linear regression model. FIG. 2 is a diagram illustrating an example of a functional configuration of a user terminal and an information processing server. FIG. 2 is a diagram illustrating an example of a hardware configuration of a user terminal and an information processing server.

In this specification, a "plan" refers to an event, campaign, promotion, etc. related to digital marketing (sales promotion using digital technology). For example, a "plan" is a sales promotion activity carried out using digital tools such as an internet-based homepage, dedicated application, or SNS (Social Networking Service), as well as digital advertising. A "plan" can promote sales of a specific product or service by specifying a specific planning period and offering benefits for the sale of the specific product or service (e.g., discounts on the sales price, points, preferential treatment for reservations and sales of other related products or services, free gifts, etc.).

The "plans" in this disclosure have the advantage of allowing for timely distribution. In other words, since there is no need to print or mail paper, information can be distributed at an overwhelmingly fast speed. For example, announcements of limited-time sales, events to be held when it rains, campaigns, promotions, etc. can be carried out in real time. In addition, it is also possible to quickly replace some of the content after distribution.

The "plans" disclosed herein have the advantage of being easier to deliver intricately targeted information compared to traditional analog sales promotion (e.g., using stickers or postcards). In other words, a method can be adopted that utilizes not only the gender, age, place of residence, etc. of the target person, but also the store visit history, content viewing history, etc. to deliver messages and advertisements. This makes it easier to deliver information to the people who need it, at the right time.

The "planning" in this disclosure has the advantage that it is possible to visualize the effects using data and analyze customers who visit the store. For example, by utilizing a function such as a digital coupon redemption function at the store, it is possible to visualize the effects of promotional activities that were difficult to achieve with traditional analog sales promotions. In addition, by accumulating data on what kind of customers visited the store and what products they purchased as a result of what promotional activities, it is possible to carry out the next promotional activity with greater precision.

Embodiments of the present disclosure will be described in detail below with reference to the attached drawings.

FIG. 1 is a diagram showing an example of a schematic configuration of an information processing device 1. The term "information processing device" may be read as an "information processing system" (the two may be synonymous). The information processing method and information processing program are realized by causing a computer (CPU: Central Processing Unit) included in the information processing device 1 to execute various processing steps.

As shown in FIG. 1, the information processing device 1 has a user terminal 10 and an information processing server 20. The user terminal 10 and the information processing server 20 are connected to each other via a network 30 so that they can communicate with each other. In FIG. 1, user terminals 10-1, 10-2, ..., 10-N (N is a natural number) are depicted as the user terminals 10, but the number of user terminals 10 is flexible and various design changes are possible. In the following, all or some of the user terminals 10-1 to 10-N may be collectively referred to as a representative symbol 10 and described. In FIG. 1, the information processing server 20 is depicted in a simplified manner as a single configuration, but this is for convenience of drawing. In reality, the information processing server 20 has a more complex configuration than that depicted in FIG. 1, both in terms of hardware configuration and software configuration. Details will be described later, but for example, the information processing server 20 may be configured to include multiple servers, databases, etc.

The information processing device (information processing system) 1 may be understood as a configuration including a user terminal 10 and an information processing server 20, or as a configuration including only the information processing server 20 without including the user terminal 10. In other words, the user terminal 10 is not a required component of the information processing device (information processing system) 1.

The user terminal 10 is, for example, a mobile terminal (mobile communication terminal) such as a mobile phone, a smartphone, or a tablet terminal. That is, the user terminal 10 in this specification is a communication device that can perform various communications with the information processing server 20. The user terminal 10 may communicate with a network (such as the Internet) via, for example, wired and/or wireless (e.g., Long Term Evolution (LTE), ^5th Generation New Radio (5G NR), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.). The communication method of the user terminal 10 has a degree of freedom, and various design changes are possible, and may be acoustic communication, visible light communication, etc. The user terminal 10 executes a browser and various other applications by user operation.

In this embodiment, the user terminal 10 can execute various operations related to "plans" that indicate events, campaigns, promotions, etc. related to digital marketing (sales promotion using digital technology) by communicating with the information processing server 20. A "plan usage application (digital sales promotion application)" for improving the user's convenience in the above-mentioned "plans" may be downloaded (installed) into the user terminal 10. By operating the user terminal 10, the user can purchase specific products or services in the above-mentioned "plans", receive benefits upon purchase, receive content distribution related to the "plans", and provide and collect the "plans" and other information.

The information processing server 20 may be any communication device capable of performing various communications with the user terminal 10, and may be interpreted as a device, circuit, etc. The information processing server 20 can control and execute various processes for providing users with suitable services using "plans" representing events, campaigns, promotions, etc. related to digital marketing (sales promotion using digital technology) via communications with the user terminal 10. In that sense, the information processing server 20 may be interpreted as a "digital marketing processing/control server, digital sales promotion processing/control server", etc.

Figure 2 is a block diagram showing an example of the internal configuration of the information processing server 20.

As shown in FIG. 2, the information processing server 20 has an identification information acquisition unit 21, an attribute information acquisition unit 22, a behavior history information acquisition unit 23, a detailed information acquisition unit 24, and a calculation mode control unit 25.

The identification information acquisition unit 21 acquires identification information of potential participants in the project. A "potential participant in the project" is defined to include at least one of the following: a person who accesses the information processing server 20 using a user terminal 10; a person who views information about the "project" using an internet-based homepage, a dedicated application, a digital tool such as a social networking service (SNS), or a digital advertisement; and a person who may purchase a product or service using the "project." When a "potential participant in the project" purchases a product or service using the "project," he or she becomes (is promoted to) a "participant in the project." A potential participant in the project may also be read as a "potential customer."

When the user terminal 10 is used to access the information processing server 20, the identification information acquisition unit 21 acquires the identification information associated with the user terminal 10 and stores this as the identification information of the potential participant of the project. The identification information of the user terminal 10 is flexible and can be modified in various ways, but can be, for example, a unique identification number such as an IMEI (International Mobile Equipment Identifier).

The attribute information acquisition unit 22 acquires attribute information of potential participants. When the user terminal 10 is used to access the information processing server 20, the attribute information acquisition unit 22 acquires attribute information including the age, sex, occupation, place of residence, hobbies, interests, etc. of the potential participants (users) of the project as user information/owner information of the user terminal 10 and/or as input information input using the user terminal 10.

The behavioral history information acquisition unit 23 acquires behavioral history information of the potential participant. When the user uses the user terminal 10 to access the information processing server 20, the behavioral history information acquisition unit 23 extracts behavioral history on a project site, etc., using the user terminal 10 as user information/owner information of the user terminal 10 and/or as input information input using the user terminal 10. The behavioral history information may include, for example, information regarding purchase history in past "projects" conducted by the same, similar, or other business operators. The purchase history information may include the content and type of purchased goods/services, the purchase time, purchase price, number of purchases, benefits received in the project as a result of purchases (e.g. cashback points), etc. Alternatively/in addition, the behavioral history information may include participation/non-participation in past projects, number of applications, and history of browsing the project site.

The detailed information acquisition unit 24 acquires detailed information about the plan. The detailed information about the plan may be interpreted as campaign information, etc., and may include the period of the plan, the menu of the plan, the format of the plan, the category of the materials (goods/services) of the plan, and the unit price of the materials (goods/services) of the plan. Alternatively/in addition, the detailed information about the plan may include whether it is linked to payment/a bonus issuance, the contents of the benefit (e.g., what is the point-back rate), etc.

The calculation mode control unit 25 controls the calculation mode of the effect of the participant candidate participating in the project. The "effect of the participant candidate participating in the project" can be calculated from various viewpoints. For example, at least one of the following is calculated as the "effect of the participant candidate participating in the project": the number or ratio of project participants for each product/service to be calculated (the number or ratio of the participant candidate who actually became a participant), the unit price of the product/service to be calculated, the number of products/services purchased by the project participants, the sales obtained by multiplying the unit price and number of the products/services to be calculated, and, if the project includes multiple products/services, the total sales obtained by adding the sales of each product/service. Alternatively, the points obtained by applying each of the above indicators (participants and sales) to a predetermined calculation formula may be calculated as the "effect of the participant candidate participating in the project". When calculating the "effect of the participant candidate participating in the project", a process is required to calculate the probability that the participant candidate will participate in the project and determine whether the participant candidate will be a "participant" or a "non-participant" based on whether the probability exceeds a predetermined threshold.

The calculation mode control unit 25 can also implement predictions of the participating purchase amount and the total number of applications in an executable manner, by adding the participation or non-participation in the project, the number of participants, and the sales from the number of participants as prediction targets. In this case, the calculation mode control unit 25 may predict the total number of applications (total number of participants) and therefore the campaign sales of purchasers who meet the specified campaign conditions for the target purchased products and purchase amount (lower limit amount). Furthermore, rather than predicting only the participation or non-participation flag for the project, the purchase amount may also be taken into account and applied to the campaign conditions to determine whether or not a person is eligible for the campaign. The calculation mode control unit 25 may also capture the purchase price per person on an ID basis and calculate the sales of the project (campaign) from the total of these.

The calculation mode control unit 25 can timely (dynamically) change the calculation mode of the "effect of a potential participant participating in a project" according to at least a part of the information acquired by the identification information acquisition unit 21, the attribute information acquisition unit 22, the behavior history information acquisition unit 23, and the detailed information acquisition unit 24, such as the type of project, the type of product/service, and the type of participants (potential participants) expected to be targeted, in addition to the situation at the time of effect calculation. In other words, the calculation mode control unit 25 holds (and can refer to) multiple calculation methods (algorithms) of the "effect of a potential participant participating in a project", and can adopt and use any of the calculation methods (algorithms) as appropriate and at the appropriate time. Below, the calculation method (algorithm) for the "effect of a potential participant participating in a project" in each mode and the control that the calculation mode control unit 25 performs in each mode will be described in detail.

The calculation mode control unit 25 calculates a first effect by applying the identification information acquired by the identification information acquisition unit 21, the attribute information acquired by the attribute information acquisition unit 22, the behavior history information acquired by the behavior history information acquisition unit 23, and the detailed information acquired by the detailed information acquisition unit 24 to a machine learning model.

The calculation mode control unit 25 calculates a second effect by fitting the detailed information acquired by the detailed information acquisition unit 24 to a linear regression model.

The calculation mode control unit 25 calculates the "effect of a potential participant participating in the project" based on a first effect based on a machine learning model and a second effect based on a linear regression model.

Effect calculation based on a machine learning model and effect calculation based on a linear regression model each have their own characteristics (merits and demerits). For example, linear regression based on a time series model calculates effects based on the figures of past projects (campaigns), so the accuracy of effect calculation such as sales and number of participants tends to be high. On the other hand, machine learning models tend to have low accuracy in predicting sales and number of participants because they are predictions on a per-user basis, but are very useful for analysis and distribution on an individual basis. In this embodiment, effect calculation based on a machine learning model and effect calculation based on a linear regression model are used selectively, and sometimes combined in an optimal form to calculate the "effect of a potential participant participating in a project," thereby making it possible to predict, verify, and utilize the effects of digital marketing.

Machine learning models, as a sort of media approach, are aimed at targeting users with a high probability of applying, and can be used to predict participation in a project (campaign) along with person-based scoring, employing a multilayer perceptron (MLP) as an analytical method. Linear regression models, as a sort of soft approach, are aimed at designing projects (campaigns) that are likely to increase the probability of application, and can analyze the causal relationship (factorial influence) between application/non-application and the project (campaign), employing multiple regression analysis or individual factor analysis as an analytical method. For example, machine learning, which can be used for output such as distribution by making predictions on an ID basis, makes it difficult to visualize the factors that "increase participation," so it is possible to use linear models in combination to interpret "what is effective in increasing participation."

Figure 3A is a diagram showing an example of detailed information acquired by the detailed information acquisition unit 24, Figure 3B is a diagram showing an example of the identification information acquired by the identification information acquisition unit 21, the attribute information acquired by the attribute information acquisition unit 22, and the behavior history information acquired by the behavior history information acquisition unit 23, and Figure 3C is a diagram showing an example of the purchase probability of a potential participant in the project calculated based on the information in Figures 3A and 3B.

In the example of Figure 3A, the following conditions are set for this plan (campaign): menu, plan format, product category, product unit price, discount rate, plan, advertising placement, and whether or not in-store measures are implemented. The following are set (obtained): retail, total price, beverage, 150 yen, 10%, yy/mm/dd-yy/mm/dd, amount/imps, and whether or not in-store measures are implemented.

In the example of FIG. 3B, age, gender, product category, and purchase history are set (acquired) for a user ID, which serves as identification information for a potential participant in the project. User ID 0001 is a 43-year-old woman, and purchase history X for product category A is set (acquired). User ID 0002 is a 25-year-old man, and purchase history Y for product category B is set (acquired). User ID 0003 is a 28-year-old woman, and purchase history Z for product category C is set (acquired).

In the example of Figure 3C, based on the information in Figures 3A and 3B, the purchase probability (promotion participation probability) of user ID 0001 is set (obtained) to 80%, the purchase probability (promotion participation probability) of user ID 0002 is set (obtained) to 70%, and the purchase probability (promotion participation probability) of user ID 0003 is set (obtained) to 20%.

Figure 4 is a diagram showing an example of effect calculation based on a machine learning model. As shown in Figure 4, the identification information acquired by the identification information acquisition unit 21 as information based on a person, the attribute information acquired by the attribute information acquisition unit 22, the behavior history information acquired by the behavior history information acquisition unit 23, and the detailed information acquired by the detailed information acquisition unit 24 as information on the project (campaign) itself are input to the machine learning model. As a result of the calculation by the machine learning model, the participation probability of the project (campaign) is output for each participating candidate (user unit) of the project. Then, based on the participation threshold (predetermined threshold), it is determined whether the participating candidate is expected to be a "participant" or a "non-participant". In other words, if the calculated participation probability is equal to or greater than the participation threshold (predetermined threshold), the participating candidate is expected to participate in the project, and if the calculated participation probability is less than the participation threshold (predetermined threshold), the participating candidate is expected not to participate in the project. Here, depending on where the participation threshold (predetermined threshold) is set, the number of participants (expected number) and sales amount for the project will fluctuate, but taking that fluctuation into account (absorbing it), the participants and sales, or the points obtained by applying these to a predecessor formula, are calculated as the "effect of potential participants taking part in the project."

The "effect of a potential participant participating in a project" obtained based on the machine learning model is called the "first effect." As will be described in more detail later, the calculation mode control unit 25 timely (dynamically) changes (corrects) the participation threshold (predetermined threshold) for calculating the "first effect."

Figure 5 shows an example of effect calculation based on a linear regression model. By clarifying the causal relationship between multiple participation factors in a project (campaign) and the resulting change in the number of applications for the project (campaign), it is possible to numerically grasp the degree of influence that each project (campaign) factor has on the number of applications for the project (campaign). Multiple regression analysis can be used with a linear regression model. Multiple regression analysis is a statistical method that, when explaining a certain result (objective variable), quantifies in the form of a function which of multiple related factors (explanatory variables) influences the result and to what extent, and expresses the relationship between the two.

In the example of Fig. 5, the number of entries (y) is set as the objective variable, while the number of winners ( _x1 ), the number of entries ( _x2 ), the number of campaign days ( _x3 ), the number of entries required for the course ( _x4 ), the number of OA PUSHes ( _x5 ), and the pseudo purchase certificate ( _x6 ) are set as explanatory variables (x). Then, the relationship between the project (campaign) factors and the number of entries for the project (campaign) (the degree of influence of both) is interpreted from the relational equation (regression equation) that represents the _objective variable with the explanatory variables and regression coefficients. Specifically, the relational equation (regression equation) y = _{β1 x1} + _{β2 x2} + _{β3 x3} + _{β4 x4} + β5 _x5 + _{β6 x6} + _β0 can be used. The explanatory variables shown in FIG. 5 are merely examples, and the number of lottery winners, the total number of points awarded, the price of the product, whether or not advertising is placed, the amount of advertising placed, etc. may be added or changed as explanatory variables.

In addition, the explanatory variables (x) illustrated in FIG. 5 are merely examples, and explanatory variables (x) may be added, deleted, or replaced as appropriate. For example, the sales of the target products in each distribution (x ₇ ) and the number of products handled in each distribution (x ₈ ) may be set as explanatory variables (x) indicating the existing sales scale in addition to the number of winners (x ₁ ), number of applied courses (x ₂ ), number of campaign days (x ₃ ), number of required courses (x _{4 ), number of OA PUSHes (x 5 )} , and pseudo purchase proof (x ₆ ). In addition, in order to reflect the potential of the target products and the existing sales scale of the distribution, the total number of products handled in the target distribution and the sales scale of the target products in the target distribution may be added. s This allows the product and distribution scale for which the plan (campaign) is implemented to be learned. Alternatively, the sales of the target products in each distribution and the number of products handled in each distribution, which indicate the existing sales scale, may be used to calculate the effect based on the machine learning model, or may be used to correct at least one of the "first effect" based on the machine learning model and the "second effect" based on the linear regression model.

The "effect of a potential participant participating in a project" obtained based on the linear regression model is called the "second effect." As will be described in more detail later, the calculation mode control unit 25 timely (dynamically) changes (corrects) the number and types of objective variables and explanatory variables used to calculate the "second effect."

In this way, the machine learning model calculates the predicted campaign participation probability for each user based on a specified threshold, generates segments based on the participation probability, and calculates the "first effect" based on a simulation of the number of campaign participants and campaign sales. The linear regression model identifies factors that have a high influence on the effect from candidate variables that have been converted into a formula, and calculates the "second effect" based on multiple regression analysis.

The "effect of the potential participant participating in the project" calculated by the calculation mode control unit 25 may be, for example, an addition and/or multiplication of parameters indicating the "first effect" from a machine learning model and the "second effect" from a linear regression model. Alternatively, the "effect of the potential participant participating in the project" may be an index obtained by substituting parameters indicating the "first effect" and the "second effect" into a specified calculation formula. Alternatively, the "effect of the potential participant participating in the project" may be a comparative display of the "first effect" and the "second effect", or may be an index obtained by substituting a parameter indicating the difference between the "first effect" and the "second effect" into a specified calculation formula.

In this way, there is a degree of freedom in how to calculate the "effect of a potential participant taking part in a project" based on the "first effect" and the "second effect," and various design modifications are possible. For example, in a machine learning model, predictions may be made by narrowing down to a population with a large amount of information in the explanatory variables, and/or in a linear regression model, a base number may be calculated and the results of the machine learning model may be added to this to obtain a predicted number. Also, the base number in the linear regression model may be summed with the numbers of high probability participants in the machine learning model, or the base number in the linear regression model may be multiplied by the participation probability lift in the machine learning model.

The calculation mode control unit 25 corrects at least one of the "first effect" based on the machine learning model and the "second effect" based on the linear regression model based on panel data surveyed by the research company. Panel data is information including questionnaires conducted by the research company (agency or its contractor), individual, awareness, purchase history, etc. Panel data is data recorded by continuously examining multiple items (e.g., project effects based on participants and sales) for the same or similar samples (e.g., the same or similar products in the same or similar projects), and is used to analyze the relationship between items over time. In a typical survey conducted continuously, the samples surveyed may differ greatly at each survey point in time, but panel data is characterized by using data continuously surveyed for the same or similar samples with almost no replacement of samples.

The calculation mode control unit 25 analyzes the panel data to predict whether the effect of the current target plan (campaign) will be up or down, and to what extent. Then, based on the analysis result of the panel data, the calculation mode control unit 25 multiplies the parameters or indicators (participants, sales, points, etc.) indicating the "first effect" and/or the parameters or indicators (participants, sales, points, etc.) indicating the "second effect" by a predetermined coefficient (e.g., a coefficient for making an upward correction such as 1.1) in accordance with the degree of the upswing prediction in the case of an upswing prediction. On the other hand, based on the analysis result of the panel data, the calculation mode control unit 25 multiplies the parameters or indicators (participants, sales, points, etc.) indicating the "first effect" and/or the parameters or indicators (participants, sales, points, etc.) indicating the "second effect" by a predetermined coefficient (e.g., a coefficient for making a downward correction such as 0.9) in accordance with the degree of the downswing prediction in the case of an upswing prediction. Here, the coefficients applied to the correction of the "first effect" and the "second effect" may be the same or different. Considering the magnitude of the influence of the panel data on the "first effect" and the "second effect," it may be preferable to use different coefficients for the correction even if the correction is the same for upside and downside.

When correcting the "second effect" based on a linear regression model, the calculation mode control unit 25 may modify the relational equation (regression equation) expressing the objective variable with explanatory variables and regression coefficients from _the basic equation, and may add or delete items set as explanatory variables, weight certain items _, etc. When the relational equation (regression equation) y = _β1x1 + _β2x2 + _β3x3 + _β4x4 + _β5x5 + _β6x6 + _β0 illustrated in Fig. ₅ is used as the basic equation, the number of applied courses ( _x2 ) and _the number _of campaign days ( _x3 ) are weighted _{, and} when it can be seen from the panel data that the importance of the pseudo purchase proof ( _x6 ) is _low , the relational equation (regression equation) may be modified to _y = _β1x1 + _2β2x2 + _{3β3x3 + β4x4 + β5x5 + β0} . That is, the coefficients multiplied by β ₂ x ₂ and β ₃ x ₃ may be increased, while β ₆ x ₆ may be excluded from the items.

When correcting the "first effect" in the machine learning model, the calculation mode control unit 25 can vary the predetermined threshold (participation threshold) used in the process of determining whether a participant candidate is expected to be a "participant" or a "non-participant" and used in comparing the probability of the participant candidate participating in the project with the probability of the participant candidate participating in the project. As an example, assume that the predetermined threshold (participation threshold) is initially set (default setting) to 70%. The calculation mode control unit 25 analyzes the panel data to predict whether the effect of the current target project (campaign) will be up or down, and to what extent. Then, based on the analysis result of the panel data, in the case of an upswing prediction, the calculation mode control unit 25 upwardly corrects the predetermined threshold (participation threshold) for determining whether a participant candidate is expected to be a "participant" or a "non-participant" (for example, 75%) according to the degree of the downswing prediction. On the other hand, based on the analysis result of the panel data, in the case of a downswing prediction, downwardly corrects the predetermined threshold (participation threshold) for determining whether a participant candidate is expected to be a "participant" or a "non-participant" (for example, 65%) according to the degree of the downswing prediction.

In this way, the calculation mode control unit 25 calculates the probability that a potential participant will participate in the project as the "effect of a potential participant participating in the project" and whether or not that probability exceeds a predetermined threshold, and corrects the predetermined threshold based on panel data surveyed by a research company. In this case, as an application of deterministic responses to project (campaign) participation, it is also possible to define individuals who participate in the project (campaign) frequently (confirmed participants) and individuals who almost never participate in the project (campaign) (confirmed non-participants), remove the confirmed participants and confirmed non-participants from the population, and determine the probability of participation for the remaining population.

The calculation mode control unit 25 may correct the "second effect (e.g., an index of participants, sales, points, etc.)" based on a linear regression model based on the "first effect (e.g., an index of participants, sales, points, etc.)" based on a machine learning model, or conversely, may correct the "first effect (e.g., an index of participants, sales, points, etc.)" based on a machine learning model based on the "second effect (e.g., an index of participants, sales, points, etc.)" based on a linear regression model. As a result, the calculation mode control unit 25 can correct the "effect of a potential participant participating in a project" based on at least one of the "first effect" based on the machine learning model and the "second effect" based on the linear regression model.

Furthermore, the calculation mode control unit 25 calculates, as the "effect of the potential participant participating in the project", the probability that the potential participant will participate in the project and whether or not that probability exceeds a predetermined threshold, and can correct the predetermined threshold based on the "first effect (e.g., an indicator of participants, sales, points, etc.)" and the "second effect (e.g., an indicator of participants, sales, points, etc.)" with at least one of them corrected. For example, if the first and second effects are corrected upward, the predetermined threshold can be corrected upward accordingly, and if the first and second effects are corrected downward, the predetermined threshold can be corrected downward accordingly.

The calculation mode control unit 25 may determine whether or not to participate based on a threshold value (or the product of probabilities) in the machine learning model, while also determining whether or not to participate on a per-user basis by feeding back and using a threshold value so that the results match the results of a linear model (after correction using panel data) that is considered to be highly accurate in terms of sales/number of participants. In this case, the threshold value may be determined by running simulations of both the machine learning model and the linear regression model for each case (project). Alternatively, multiple expected conditions may be simulated in advance, and the threshold value may be determined in advance to minimize error.

The calculation mode control unit 25 can perform parameter correction (e.g., correction of a threshold value for determining the participation probability described above) for calculating the first effect based on the machine learning model based on the second effect (calculation result) based on the linear regression model. The calculation mode control unit 25 can perform parameter correction (e.g., correction of various parameters and coefficients (input parameters and output parameters) of the linear regression relational equation described above) for calculating the second effect based on the linear regression model (output value of the linear regression relational equation) based on the first effect (calculation result) based on the machine learning model.

Based on the first effect based on the machine learning model and the second effect based on the linear regression model (existing calculation results), the calculation mode control unit 25 can also perform parameter correction (e.g., correction of a threshold value for determining the participation probability described above) to calculate the first effect based on the machine learning model, and/or parameter correction (e.g., correction of various parameters and coefficients (input parameters and output parameters) of the linear regression equation described above) to calculate the second effect based on the linear regression model (output value of the linear regression equation).

When predicting the "predetermined threshold" that serves as an index (criteria) for determining whether a participant candidate is expected to be a "participant" or a "non-participant," the calculation mode control unit 25 may refer to a specialized model for each project (campaign) that is separate (independent) from the machine learning model and the linear regression model. This specialized model is, so to speak, a specialized algorithm for predicting a predetermined threshold, and may be, for example, a model that directly predicts the threshold from campaign information and data on the number of resulting participants, rather than on a per-person basis. The specialized model may be created and stored by the information processing server 20, or may be created by an external device of the information processing server 20 and introduced into the information processing server 20. In addition, the specialized model may be updated (upgraded) as appropriate depending on the content and progress of the project (campaign).

Furthermore, the "predetermined threshold" which is the index (criteria) for determining whether a potential participant is expected to be a "participant" or a "non-participant" may be set to a different value for each business operator (each platform), or may be set to a common value for multiple business operators (platforms). In the former case, there is an advantage that the effects of digital marketing can be predicted, verified, and utilized in a manner that is in line with the tastes, preferences, and characteristics of the business operator (platform). In the latter case, there is an advantage that the effects of digital marketing can be predicted, verified, and utilized in a manner that is mutually comparable using evaluation criteria common to multiple business operators (platforms). In addition to the above "predetermined threshold", various parameters (campaign variables, so to speak) required for obtaining the effects of digital marketing (first and second effect calculations) may also be set to a different value for each business operator (each platform), or may be set to a common value for multiple business operators (platforms).

In addition, when obtaining the effect of digital marketing (first and second effect calculations), the period of the target plan (campaign) may be divided into multiple periods, and the effect of digital marketing may be obtained (first and second effect calculations) for each divided period. For example, the period of a plan (campaign) on a monthly basis may be divided into weekly/daily periods, and the effect of digital marketing may be obtained (first and second effect calculations) on a weekly/daily basis. In order to compensate for the small number of plans (campaigns) (number of samples), participants (potential participants) may be divided into weekly/daily periods, and the disadvantages of dividing the period may be eliminated by considering the progress variable during the plan (campaign) period (for example, a progress variable such as a percentage of days has passed, taking into account initial speed and last-minute participation, etc.). For example, if the initial speed is strong, the progress variable near 0% will be strong, and if the last-minute participation is strong, the progress variable near 100% will be strong.

The calculation mode control unit 25 may set multiple rules to prevent abnormal predictions of the "first effect," the "second effect," and the "effect of the participation of the candidate in the project." For example, for the category market size, since the demand for the category does not change suddenly, the market size may be determined by taking into account the purchases of the entire category, including competitors, and the trend of the category over time. For the category share, the share of the competing brand may be determined as the maximum/minimum share in the past (for example, the past three years of major brands). For the company's purchasing seasonality, the maximum/minimum purchases of the company's brand in that season may be determined taking seasonality into consideration. Here, the category share and company's purchasing seasonality may be allowed to be within a range that does not become an outlier (for example, 1.5 times the interquartile range) assuming that an unprecedented change occurs. Furthermore, the range of the category share and company's purchasing seasonality linked to the category market size may be capped as an abnormal value and numerical correction may be performed.

In the information processing device (information processing system) 1, a simulation model may be constructed both inside and outside the platform environment. For example, outside the platform environment, the aggregated results (e.g., the number of campaign participants predicted from campaign information, product information, and incentives) may be obtained using a simple simulation model and stored and output. On the other hand, within the platform environment, an ad-hoc simulation may be performed in the platform environment to predict whether or not participants will participate for each ID based on campaign information, product information, and incentives, thereby making it possible to predict the number of participants and use the results for targeted delivery, etc.

A simple simulation has the advantage of being able to produce results in real time, but the disadvantage is that it is less accurate than an ad hoc simulation. An ad hoc simulation has the advantage of being more accurate than a simple simulation, but the disadvantage is that it takes time to produce results because manual work is added. One way to use both a simple simulation and an ad hoc simulation is, for example, to enter the simulation conditions, then run the simple simulation and view the results, and when more precise results are required, to request the execution of an ad hoc simulation, run the ad hoc simulation, and view the results.

In this way, in the information processing device according to the present disclosure, the identification information acquisition unit acquires identification information of the potential participants in the project, the attribute information acquisition unit acquires attribute information of the potential participants, the behavioral history information acquisition unit acquires behavioral history information of the potential participants, the detailed information acquisition unit acquires detailed information about the project, and the calculation mode control unit controls the calculation mode of the effect of the potential participants participating in the project. This makes it possible to optimally predict, verify, and utilize the effects of digital marketing.

The information processing device according to the present disclosure can be realized by developing a simulation tool that is specialized for platform promotion and predicts the probability of participation on an ID (identification information of a potential participant in a project) basis. The information processing device according to the present disclosure achieves both readability (interpretation of factors for participation) and generalizability, and by combining and utilizing individual model selections, can predict, verify, and utilize the effects of digital marketing with high accuracy. Furthermore, in order to prevent abnormal values from being output as the effects of digital marketing, it is possible to further improve the accuracy of predicting, verifying, and utilizing the effects of digital marketing by setting upper and lower limits based on panel data surveyed by a research company or by adding ingenuity to rule processing outside the machine learning model.

Conventional simulation models have generally been prediction models based on time series data, not on ID (identification information of potential participants in a project) units. This is because predictions on an ID-by-ID basis often have limited accuracy. However, for this reason, conventional simulation models are merely simulations and cannot be used for actual marketing or for differentiating ad delivery. In one aspect of this embodiment, while requiring effect calculation on an ID (identification information of potential participants in a project) basis, a certain level of quality is guaranteed in terms of accuracy, and a processing mechanism is incorporated that prevents abnormal values from being output as the effect of digital marketing in rule processing outside the machine learning model. As a result, a prediction model based on ID (identification information of potential participants in a project) units can be realized while maintaining a certain level of accuracy and quality. Furthermore, in order to achieve both the readability (meaning) and generalizability (accuracy) of the model, both linear regression and machine learning models are implemented and reflected in the output results.

The information processing device according to the present disclosure can predict the effects and be applied to the design and distribution of plans (campaigns) by analyzing factors that affect the results of the plan (campaign). By predicting the effects of the plan (campaign), it is possible to provide businesses with a convincing case and reduce the amount of work required for simulations and persuasion. It is also possible to provide businesses with highly original optimization of digital marketing. It is also possible to increase the effectiveness of the plan (campaign) by analyzing effective elements on a user ID basis, and to improve the effectiveness of direct distribution. For businesses, it is also possible to use this as a basis for budget allocation by performing budget and optimization allocation for mass areas.

The information processing device according to the present disclosure builds an algorithm that predicts the effects of digital sales promotion using machine learning models and linear regression models from past campaign data such as open information and data held by platform operators. When information on planned sales promotion measures (e.g., number of points, campaign period, target distribution, average number of purchases, etc.) is entered, the "campaign effects such as number of participants and sales" are automatically calculated. In addition, it is also possible to derive campaign conditions that maximize the "campaign effects such as number of participants and sales," thereby realizing optimization of digital sales promotion. Furthermore, it is possible to build a model that is uniquely customized for each digital sales promotion operator.

Figure 6 is a diagram showing an example of the functional configuration of the user terminal 10 and the information processing server 20. As shown in Figure 6, the user terminal 10 and the information processing server 20 have a control unit 110, a storage unit 120, a communication unit 130, an input unit 140, and an output unit 150. Note that Figure 6 shows an example of the functional blocks of the user terminal 10 and the information processing server 20, and they may have other functional blocks not shown in Figure 6. In addition, they may be configured not to include some of the functional blocks.

The control unit 110 controls the user terminal 10 and the information processing server 20. The control unit 110 can be configured as a controller, a control circuit, or a control device that is described based on a common understanding in the technical field related to the present invention.

The memory unit 120 stores information used by the user terminal 10 and the information processing server 20. The memory unit 120 can be configured, for example, with a memory, storage, storage device, etc. that are described based on a common understanding in the technical field related to the present invention.

The communication unit 130 performs communication between the user terminal 10 and the information processing server 20 (e.g., mutual communication between the two). The communication unit 130 can be configured with a transmitter/receiver, a transmission/reception circuit, or a transmission/reception device that is described based on a common understanding in the technical field related to the present invention. The communication unit 130 may be configured with a transmission unit and a reception unit.

The input unit 140 accepts inputs at the user terminal 10 and the information processing server 20. The input unit 140 may also be connected to a specific device, storage medium, etc., and accept data input. The input unit 140 may output the input results to, for example, the control unit 110. The input unit 140 may be configured with input devices such as a keyboard, mouse, and buttons, input/output terminals, input/output circuits, etc., which are described based on common understanding in the technical field related to the present invention. The input unit 140 may also be configured as an integrated unit with a display unit (for example, a touch panel).

The output unit 150 performs output on the user terminal 10 and the information processing server 20. For example, the output unit 150 may be configured to include a display unit that displays images, an audio output unit that outputs audio, and the like. The display unit may be configured, for example, by a display device such as a display or monitor that is described based on a common understanding in the technical field related to the present invention. Furthermore, the audio output unit may be configured by an output device such as a speaker that is described based on a common understanding in the technical field related to the present invention.

The block diagrams used to explain the above embodiments show functional blocks. These functional blocks (components) are realized by any combination of hardware and/or software. Furthermore, the means of realizing each functional block is not particularly limited. That is, each functional block may be realized by a single device that is physically connected, or may be realized by two or more physically separate devices that are connected by wire or wirelessly.

For example, the devices (user terminal 10 and information processing server 20) in one embodiment of the present disclosure may function as a computer that performs the information processing (digital marketing processing, digital sales promotion processing) of the present disclosure. FIG. 7 is a diagram showing an example of the hardware configuration of the user terminal 10 and the information processing server 20. The above-mentioned user terminal 10 and information processing server 20 may be physically configured as a computer device including a processor 1001, memory 1002, storage 1003, communication device 1004, input device 1005, output device 1006, bus 1007, etc.

In this disclosure, the terms apparatus, circuit, device, unit, server, etc. may be used interchangeably. The hardware configurations of the user terminal 10 and the information processing server 20 may be configured to include one or more of the devices shown in the figures, or may be configured to exclude some of the devices.

For example, although only one processor 1001 is shown, there may be multiple processors. Furthermore, processing may be performed by one processor, or processing may be performed by two or more processors simultaneously, sequentially, or using other techniques. Furthermore, the processor 1001 may be implemented by one or more chips.

The functions of the user terminal 10 and the information processing server 20 are realized by loading specific software (programs) onto hardware such as the processor 1001 and memory 1002, causing the processor 1001 to perform calculations and control communication via the communication device 1004, and the reading and/or writing of data in the memory 1002 and storage 1003, etc.

The processor 1001, for example, operates an operating system to control the entire computer. The processor 1001 may be configured as a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic unit, a register, etc. Each unit, such as the control unit 110 described above, may be realized by the processor 1001.

The processor 1001 also reads out programs (program codes), software modules, data, etc. from at least one of the storage 1003 and the communication device 1004 into the memory 1002, and executes various processes according to these. The programs used are those that cause a computer to execute at least some of the operations described in the above-mentioned embodiments. For example, the control unit 110 may be realized by a control program stored in the memory 1002 and running on the processor 1001, and similar implementations may be made for other functional blocks.

The memory 1002 is a computer-readable recording medium, and may be composed of, for example, at least one of a ROM (Read Only Memory), an EPROM (Erasable Programmable ROM), an EEPROM (Electrically EPROM), a RAM (Random Access Memory), or other suitable storage medium. The memory 1002 may also be called a register, a cache, a main memory, or the like. The memory 1002 may store executable programs (program codes), software modules, and the like for implementing a method according to one embodiment.

Storage 1003 is a computer-readable recording medium, and may be composed of at least one of a flexible disk, a floppy disk, a magneto-optical disk (e.g., a compact disk (CD-ROM (Compact Disc ROM)), a digital versatile disk, a Blu-ray (registered trademark) disk), a removable disk, a hard disk drive, a smart card, a flash memory device (e.g., a card, a stick, a key drive), a magnetic stripe, a database, a server, and other suitable storage media. Storage 1003 may be called an auxiliary storage device. Note that the above-mentioned storage unit 120 may be realized by memory 1002 and/or storage 1003.

The communication device 1004 is hardware (transmission/reception device) for communicating between computers via at least one of a wired network and a wireless network, and is also called, for example, a network device, a network controller, a network card, or a communication module. The communication device 1004 may include a SIM card. The above-mentioned communication unit 130 may be realized by the communication device 1004.

The input device 1005 is an input device (e.g., a keyboard, a mouse, etc.) that accepts input from the outside. The output device 1006 is an output device (e.g., a display, a speaker, etc.) that performs output to the outside. The input device 1005 and the output device 1006 may be integrated into one (e.g., a touch panel). The above-mentioned input unit 140 and output unit 150 may be realized by the input device 1005 and the output device 1006, respectively.

In addition, each device such as the processor 1001 and memory 1002 is connected by a bus 1008 for communicating information. The bus 1007 may be configured as a single bus, or may be configured with different buses between the devices.

The user terminal 10 and the information processing server 20 may also be configured to include hardware such as a microprocessor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a programmable logic device (PLD), or a field programmable gate array (FPGA), and some or all of the functional blocks may be realized by the hardware. For example, the processor 1001 may be implemented by at least one of these pieces of hardware.

(Modification)
In addition, terms explained in this disclosure and/or terms necessary for understanding this disclosure may be replaced with terms having the same or similar meanings.

The information, parameters, etc. described in this disclosure may be expressed using absolute values, may be expressed using relative values from a predetermined value, or may be expressed using other corresponding information. Furthermore, the names used for parameters, etc. in this disclosure are not limiting in any way.

The information, signals, etc. described in this disclosure may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, optical fields or photons, or any combination thereof.

Information, signals, etc. may be input and output via multiple network nodes. The input and output information, signals, etc. may be stored in a specific location (e.g., memory) or may be managed using a table. The input and output information, signals, etc. may be overwritten, updated, or added to. Output information, signals, etc. may be deleted. Input information, signals, etc. may be transmitted to another device.

In addition, notification of specific information (e.g., notification that "X is the case") is not limited to explicit notification, but may be implicit (e.g., by not notifying the specific information or by notifying other information).

Software shall be construed broadly to mean instructions, instruction sets, code, code segments, program code, programs, subprograms, software modules, applications, software applications, software packages, routines, subroutines, objects, executable files, threads of execution, procedures, functions, etc., whether referred to as software, firmware, middleware, microcode, hardware description language, or otherwise.

Software, instructions, information, etc. may also be transmitted and received via at least one of a transmission medium and a signal waveform. For example, if software is transmitted from a website, server, or other remote source using at least one of wired technologies (such as coaxial cable, fiber optic cable, twisted pair, Digital Subscriber Line (DSL)) and wireless technologies (such as infrared, microwave), then at least one of these wired and wireless technologies is included within the definition of a transmission medium.

As used in this disclosure, the terms "system" and "network" may be used interchangeably.

Each aspect/embodiment described in this disclosure may be used alone, in combination, or switched between depending on the implementation. In addition, the processing procedures, sequences, flow charts, etc. of each aspect/embodiment described in this disclosure may be rearranged as long as there is no inconsistency. For example, the methods described in this disclosure present elements of various steps using an exemplary order, and are not limited to the particular order presented.

As used in this disclosure, the phrase "based on" does not mean "based only on," unless expressly stated otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

Any reference to elements using designations such as "first," "second," etc., used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used in this disclosure as a convenient way to distinguish between two or more elements. Thus, a reference to a first and a second element does not imply that only two elements may be employed or that the first element must precede the second element in some way.

When the terms "include," "including," and variations thereof are used in this disclosure, these terms are intended to be inclusive, similar to the term "comprising." Additionally, the term "or," as used in this disclosure, is not intended to be an exclusive or.

In this disclosure, where articles have been added through translation, such as a, an, and the in English, this disclosure may include that the nouns following these articles are in the plural form.

Although the invention disclosed herein has been described in detail above, it is clear to those skilled in the art that the invention disclosed herein is not limited to the embodiments described herein. The invention disclosed herein can be implemented in modified and altered forms without departing from the spirit and scope of the invention as defined by the claims. Therefore, the description of the disclosure is intended as an illustrative example and does not impose any limiting meaning on the invention disclosed herein.

1. Information processing device (information processing system)
10 (10-1, 10-2, ..., 10-N) User terminal 20 Information processing server (digital marketing processing/control server, digital sales promotion processing/control server)
21: Identification information acquisition unit 22: Attribute information acquisition unit 23: Action history information acquisition unit 24: Detailed information acquisition unit 25: Calculation mode control unit 30: Network

Claims (8)

an identification information acquisition unit that acquires identification information of a potential participant of the project;
An attribute information acquisition unit that acquires attribute information of the participant candidates;
A behavior history information acquisition unit that acquires behavior history information of the participant candidate;
A detailed information acquisition unit that acquires detailed information about the plan;
a calculation mode control unit that controls a calculation mode of an effect of the participation of the participant candidate in the project;
having
the calculation mode control unit calculates the effect of the potential participant participating in the project based on a first effect obtained by applying the identification information, the attribute information, the action history information, and the detailed information to a machine learning model, and a second effect obtained by applying the detailed information to a linear regression model;
23. An information processing apparatus comprising: The calculation mode control unit corrects at least one of the first effect and the second effect based on panel data surveyed by a research company.
2. The information processing apparatus according to claim 1 , the calculation mode control unit calculates, as the effect of the participation candidate's participation in the event, a probability that the participation candidate will participate in the event, whether or not the probability exceeds a predetermined threshold, and an output value of a linear regression relational equation, and corrects the predetermined threshold and a parameter of the linear regression relational equation based on the first effect and the second effect, at least one of which has been corrected.
3. The information processing apparatus according to claim 2 . an identification information acquisition unit that acquires identification information of a potential participant of the project;
An attribute information acquisition unit that acquires attribute information of the participant candidates;
A behavior history information acquisition unit that acquires behavior history information of the participant candidate;
A detailed information acquisition unit that acquires detailed information about the plan;
a calculation mode control unit that controls a calculation mode of an effect of the participation of the participant candidate in the project;
having
the calculation mode control unit calculates, as the effect of the participation of the participant candidate in the project, a probability that the participant candidate will participate in the project, whether or not the probability exceeds a predetermined threshold, and an output value of a linear regression relational equation, and corrects the predetermined threshold and a parameter of the linear regression relational equation based on panel data obtained by a survey conducted by a research company.
23. An information processing apparatus comprising: acquiring identification information of a potential participant of the project by an identification information acquisition unit;
acquiring attribute information of the participant candidates by an attribute information acquiring unit;
A step of acquiring behavior history information of the participant candidate by a behavior history information acquisition unit;
A step of acquiring detailed information about the plan by a detailed information acquisition unit;
A step of controlling a calculation mode control unit to calculate a calculation mode of an effect of the participation of the participant candidate in the project;
having
In the controlling step, the calculation mode control unit calculates the effect of the potential participant participating in the project based on a first effect obtained by applying the identification information, the attribute information, the action history information, and the detailed information to a machine learning model, and a second effect obtained by applying the detailed information to a linear regression model.
An information processing method characterized in that each of the above processing steps is executed by a computer. acquiring identification information of a potential participant of the project by an identification information acquisition unit;
acquiring attribute information of the participant candidates by an attribute information acquiring unit;
A step of acquiring behavior history information of the participant candidate by a behavior history information acquisition unit;
A step of acquiring detailed information about the plan by a detailed information acquisition unit;
A step of controlling a calculation mode control unit to calculate a calculation mode of an effect of the participation of the participant candidate in the project;
having
In the controlling step, the calculation mode control unit calculates, as the effect of the participation of the participant candidate in the project, a probability that the participant candidate will participate in the project, whether or not the probability exceeds a predetermined threshold, and an output value of a linear regression relational equation, and corrects the predetermined threshold and a parameter of the linear regression relational equation based on panel data obtained by a survey conducted by a research company.
An information processing method characterized in that each of the above processing steps is executed by a computer. acquiring identification information of a potential participant of the project by an identification information acquisition unit;
acquiring attribute information of the participant candidates by an attribute information acquiring unit;
A step of acquiring behavior history information of the participant candidate by a behavior history information acquisition unit;
A step of acquiring detailed information about the plan by a detailed information acquisition unit;
A step of controlling a calculation mode control unit to calculate a calculation mode of an effect of the participation of the participant candidate in the project;
having
In the controlling step, the calculation mode control unit calculates the effect of the potential participant participating in the project based on a first effect obtained by applying the identification information, the attribute information, the action history information, and the detailed information to a machine learning model, and a second effect obtained by applying the detailed information to a linear regression model.
An information processing program that causes a computer to execute each of the above processing steps . acquiring identification information of a potential participant of the project by an identification information acquisition unit;
acquiring attribute information of the participant candidates by an attribute information acquiring unit;
A step of acquiring behavior history information of the participant candidate by a behavior history information acquisition unit;
A step of acquiring detailed information about the plan by a detailed information acquisition unit;
A step of controlling a calculation mode control unit to calculate a calculation mode of an effect of the participation of the participant candidate in the project;
having
In the controlling step, the calculation mode control unit calculates, as the effect of the participation of the participant candidate in the project, a probability that the participant candidate will participate in the project, whether or not the probability exceeds a predetermined threshold, and an output value of a linear regression relational equation, and corrects the predetermined threshold and a parameter of the linear regression relational equation based on panel data obtained by a survey conducted by a research company.
An information processing program that causes a computer to execute each of the above processing steps .

Images