Information processing device, information processing method, and information processing program

Abstract

This technology provides a method for calculating an uplift score that represents purchasing trends. [Solution] The information processing device estimates a group of users consisting of multiple users into a first user group that has undergone an intervention to induce a predetermined conversion, a first probability that a user who has achieved the conversion is included in the first user group, a second probability that a user who has achieved the conversion is included in the second user group, a third probability that the entire user group has achieved the conversion, and uses the first, second, and third probabilities to calculate an uplift score that represents the effect of the intervention on the user group.

Description

Technical Field

[0001] The present invention relates to a technology for calculating uplift scores.

Background Art

[0002] In the field of advertising, it is a difficult problem to determine whether user purchases after implementing a predetermined intervention (for example, implementing a marketing measure such as distributing coupons) can truly be considered an effect of the intervention. On the other hand, if such knowledge can be obtained, marketing activities can be made more efficient. For example, when distributing advertisements containing coupons online, it is possible to select users who will make a purchase when receiving a coupon as targets, and distribute advertisements only to those targets. As described above, by narrowing the target distribution of advertisements to those who can be expected to respond to the advertisement's effect, it is expected that marketing activities can be made effective and efficient.

[0003] As a method for selecting targets that can be expected to improve the effect of an intervention as described above, uplift modeling is known. Uplift modeling is a method for estimating which users should be targeted for intervention in order to improve intervention effects. In uplift modeling, an uplift score, which is an index for target selection, is calculated, and target estimation is performed based on the calculated uplift score.

[0004] The uplift score can be calculated, for example, as described in Patent Document 1, based on the difference between the probability of purchase when an intervention is implemented and the probability of purchase when no intervention is implemented. Since a higher uplift score indicates a higher intervention effect, selecting a group of users with higher uplift scores as targets makes it possible to carry out marketing activities that can be expected to achieve high intervention effects.

Prior Art Literature

Patent Literature

[0005] [Patent Document 1] Japanese Patent Publication No. 2024-131197 [Non-patent literature]

[0006] [Non-Patent Document 1] Dmitri Goldenberg et al., “Free Lunch! Retrospective Uplift Modeling for Dynamic Promotions Recommendation within ROI Constraints”, August 2020, arXiv: 2008.06293. [Overview of the project] [Problems that the invention aims to solve]

[0007] As disclosed in Patent Document 1, according to the conventional uplift calculation method, the uplift score can be calculated as the difference between the probability of purchase when an intervention is performed and the probability of purchase when no intervention is performed. However, this calculation method does not take into account the probability that all users actually made a purchase (i.e., the probability of conversion). By taking into account the probability that all users made a purchase, the uplift score, which is an indicator for target selection, can become an indicator that better represents the purchase tendency.

[0008] In light of the above issues, this disclosure aims to provide a technology for calculating an uplift score that represents purchasing trends. [Means for solving the problem]

[0009] To solve the above problems, one aspect of the information processing apparatus according to the present invention includes: a grouping unit that groups a user group consisting of multiple users into a first user group that has undergone an intervention to induce a predetermined conversion and a second user group that has not undergone the intervention; a first estimation unit that estimates a first probability, which is the probability that a user who has achieved the conversion is included in the first user group, and a second probability, which is the probability that a user who has achieved the conversion is included in the second user group; a second estimation unit that estimates a third probability, which is the probability that the entire user group has achieved the conversion; and a calculation unit that uses the first, second, and third probabilities to calculate an uplift score that represents the effect of the intervention on the user group.

[0010] To solve the above problems, one aspect of the information processing method according to the present invention includes: grouping a user group consisting of multiple users into a first user group that has undergone an intervention to induce a predetermined conversion and a second user group that has not undergone the intervention; estimating a first probability, which is the probability that a user who has achieved the conversion is included in the first user group, and a second probability, which is the probability that a user who has achieved the conversion is included in the second user group; estimating a third probability, which is the probability that the entire user group has achieved the conversion; and calculating an uplift score, which represents the effect of the intervention on the user group, using the first, second, and third probabilities. include.

[0011] To solve the above problems, one aspect of the information processing program according to the present invention is an information processing program for causing a computer to perform information processing, the program for causing the computer to perform the following processes: a grouping process that groups a user group consisting of multiple users into a first user group that has undergone intervention to induce a predetermined conversion and a second user group that has not undergone the intervention; a first estimation process that estimates a first probability that the user who achieved the conversion is included in the first user group and a second probability that the user who achieved the conversion is included in the second user group; a second estimation process that estimates a third probability that the entire user group achieved the conversion; and a calculation process that uses the first probability, the second probability, and the third probability to calculate an uplift score that represents the effect of the intervention on the user group. [Effects of the Invention]

[0012] According to the present invention, a technique is provided for calculating an uplift score that represents purchasing trends. The objects, embodiments, and effects of the present invention described above, as well as any other objects, embodiments, and effects of the present invention not described above, can be understood by those skilled in the art from the following embodiments for carrying out the invention by referring to the accompanying drawings and the claims. [Brief explanation of the drawing]

[0013] [Figure 1] Figure 1 shows an example of the configuration of an information processing system according to one embodiment. [Figure 2] Figure 2 shows a conceptual diagram of the uplift score calculation process according to one embodiment. [Figure 3] Figure 3 shows an example of the functional configuration of an e-commerce server according to one embodiment. [Figure 4] Figure 4 shows an example of the functional configuration of an information processing device according to one embodiment. [Figure 5] Figure 5 shows an example of the hardware configuration of an information processing device according to one embodiment. [Figure 6]Figure 6 shows a flowchart of the process performed by an information processing system according to one embodiment. [Modes for carrying out the invention]

[0014] Hereinafter, embodiments 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 function are denoted by the same reference numeral, and their descriptions are omitted. The embodiments disclosed below are merely examples of means for realizing the present invention, and should be modified or changed as appropriate depending on the configuration of the apparatus to which the present invention is applied and various conditions, and the present invention is not limited to the embodiments below. Furthermore, not all combinations of features described in these embodiments are essential for solving the problem of the present invention.

[0015] [Configuration of the Information Processing System] Figure 1 shows an example of the configuration of an information processing system 1 according to this embodiment. The information processing system 1 comprises an information processing device 10, an e-commerce server 11, and a user device 12. The information processing device 10, the e-commerce server 11, and the user device 12 are configured to communicate with each other via a network 13. The network 13 can include the Internet, an intranet, a LAN (Local Area Network), a WAN (Wide Area Network), a mobile communication network, etc. Although Figure 1 shows one user device 12, the information processing system 1 is configured to have multiple user devices having similar functions to user device 12, and in this disclosure, these multiple user devices are collectively referred to as user device 12. Furthermore, user device 12 is operated by a user 14. In this disclosure, the terms user device and user may be understood as synonymous.

[0016] The electronic commerce server 11 is a server device capable of providing services for electronic commerce. In the present disclosure, the electronic commerce server 11 will be described as a server device that operates an E-commerce mall (mall-type E-commerce site) in which a plurality of stores develop a shopping mall online. Specifically, the electronic commerce server 11 operates an E-commerce mall that develops a shopping mall via sales pages (web pages) for products sold by a plurality of stores (merchants). In the E-commerce mall, items including tangible and / or intangible products and services are sold. In the following description, the explanation will be given targeting products, but the same explanation can be applied to other items as well.

[0017] The electronic commerce server 11 receives access from a user 14 via a network 13 from a user device 12, and can provide various services related to shopping in the E-commerce mall to the user 14. For example, in response to the user 14 accessing the E-commerce mall and performing an action such as purchase or click on a sales page of an arbitrary product, the electronic commerce server 11 provides a service related to the product to the user 14. In addition, the electronic commerce server 11 can acquire (collect) and manage information related to clicks and purchases made by the user 14 on sales pages of products sold in the E-commerce mall. In the present disclosure, an action such as purchase of an item provided in a service for electronic commerce provided by the electronic commerce server 11, which is an expected action for a user using the service, is referred to as a conversion. For example, a conversion is an action such as a click or purchase on a sales page by the user 14, and includes an action that the operator of the electronic commerce server 11 expects the user 14 to perform in the E-commerce mall.

[0018] The electronic commerce server 11 can perform any intervention to encourage conversion for a user who uses a service for electronic commerce. For example, the electronic commerce server 11 performs intervention by distributing an advertisement related to the service for electronic commerce (an E-commerce mall in the present embodiment). The advertisement may include information such as coupons and discount tickets that can be used in the service. The electronic commerce server 11 may distribute the advertisement, for example, by means of a banner or the like in the E-commerce mall accessed by the user, or may distribute or provide the advertisement by e-mail or other means.

[0019] In the present disclosure, a user who is subjected to arbitrary (for example, predetermined) intervention is also referred to as an "intervention user (treated user)". On the other hand, in the present disclosure, not performing a predetermined intervention despite controlling the service or the like is referred to as "controlling (control)", and a user who is not subjected to any intervention is also referred to as a "control user (control user)". As described later, the electronic commerce server 11 can determine which user is set as an intervention user based on an uplift score calculated by the information processing apparatus 10.

[0020] The user 14 can operate the user apparatus 12 to access the electronic commerce server 11 and receive a service provided by the electronic commerce server 11. In the present disclosure, the user 14 can receive various services in the E-commerce mall provided by the electronic commerce server 11. For example, the user 14 can operate the user apparatus 12 to access the E-commerce mall, browse sales pages of various products provided in the E-commerce mall, and purchase products sold on the sales pages. The user 14 can purchase the product by adding the product to a shopping cart on the product sales page, inputting predetermined information and performing payment.

[0021] User 14 registers information about themselves (hereinafter also referred to as user attributes) in order to use the services in the e-commerce mall. In one embodiment, User 14 sets up an account with a user ID (user identification information) that identifies User 14 and is linked to User 14's user attributes. Then, User 14 logs into the e-commerce mall provided by the e-commerce server 11 using User 14's account and uses the services in the e-commerce mall. By setting up a user ID, User 14 can use the services of the e-commerce mall even from a user device other than the user device 12 connected to the network 13. The types and number of user attributes that User 14 should register may be predetermined by the e-commerce server 11.

[0022] User attributes will be explained using the example of user attributes for user 14. User attributes for user 14 may include user 14's address, user 14's name, and user 14's demographic information (such as gender, age, residential area, occupation, family structure, and other demographic information about user 14). User attributes for user 14 may also include the registration number and registered name when using web services, including the e-commerce mall provided by the e-commerce server 11. Furthermore, user attributes for user 14 may include information about the usage history, search history, product purchase history, and points that can be accumulated through the use of web services, including the e-commerce mall, provided by the e-commerce server 11. Thus, user attributes for user 14 can include any attribute information, including information related to user 14 themselves and information about the usage history of web services, including the e-commerce mall. Such attributes related to user 14's registration and actual usage history of web services can also be called user 14's factual user attributes.

[0023] User 14's user attributes may include not only factual user attributes but also estimated user attributes. These estimated user attributes can be estimated based on user 14's factual user attributes, for example, by a trained user attribute estimation model. The estimated user attributes may include products that user 14 is estimated to be interested in, their preferences, and / or the lifestyle estimated for user 14.

[0024] The user device 12 is an information processing device such as a smartphone, mobile phone, PDA (Personal Digital Assistant), or tablet terminal. The user device 12 is configured to communicate with the e-commerce server 11 and the information processing device 10 via the network 13. The user device 12 has a display unit (display surface) such as an LCD display, and the user 14 can perform various operations using the GUI (Graphical User Interface) equipped on the display unit. These operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or stylus. The user device 12 may have a separate display unit and may be a notebook PC (Personal Computer) or a desktop PC.

[0025] The e-commerce server 11 can obtain user attributes of users 14 who access the e-commerce mall it provides, associating them with the user IDs of the users 14. The e-commerce server 11 can also obtain conversion information by users 14 (for example, conversion history with time information and history of whether or not conversions occurred over a certain period), associating it with the user IDs. The e-commerce server 11 provides the user attributes and conversion information associated with the user IDs to the information processing device 10.

[0026] The information processing device 10 calculates an uplift score based on user attributes and conversion information associated with the user ID, obtained from the e-commerce server 11. The uplift score corresponds to an indicator for selecting targets to improve the effectiveness of the intervention. The procedure for calculating the uplift score according to this embodiment will be described later. In Figure 1, the information processing device 10 and the e-commerce server 11 are configured as separate devices, but the two devices may be configured as a single device, for example, the information processing device 10 may be configured to include the functions of the e-commerce server 11.

[0027] [How to calculate the uplift score] The method for calculating the uplift score is described below. In this disclosure, the uplift score is calculated for each user group, which is a group of one or more users having one or more common (identical) user attributes. In this disclosure, one or more user groups having one or more common user attributes are referred to as user groups having user feature = X. When different user attributes are represented as x1, x2, ..., user feature = X may be represented as a vector of one or more user attributes (x1, x2, ...). For explanatory purposes, the presence or absence of intervention is represented by the variable t, with t=1 representing intervention and t=0 representing no intervention. Furthermore, the user group with intervention (t=1) is referred to as the intervention user group, and the user group without intervention (t=0) is referred to as the control user group. Furthermore, the presence or absence of conversion (i.e., whether or not a conversion occurred) is represented by the variable y, with y=1 representing conversion and y=0 representing no conversion.

[0028] (A) Conventional method for calculating the uplift score First, let's explain the conventional method for calculating the uplift score. The probability of conversion (y=1) occurring among the intervention user group (t=1) of the user group with user characteristics = X is expressed as p(y=1|t=1,X). Similarly, the probability of conversion (y=1) occurring among the control user group (t=0) of the user group with user attributes = X is expressed as p(y=1|t=0,X). The conventional uplift score can be calculated using the difference between these two probabilities, as shown in equation (1).

number

[0029] The probabilities p(y=1|t=1,X) and p(y=1|t=0,X) can be estimated by machine learning. For this purpose, the information processing device 10 may, for example, use a certain amount of data from the user attributes and conversion information data associated with the user ID obtained from the e-commerce server 11 for training, and use the remaining data for estimation (prediction).

[0030] To estimate probabilities p(y=1|t=1,X) and p(y=1|t=0,X) and calculate the uplift score, one example of an approach can be used: a single ML (machine learning) model called S-Learner, two separate ML models called T-Learner, or a class transformation approach. Below, we will describe (A-1) S-Learner, (A-2) T-Learner, and (A-3) class transformation.

[0031] (A-1) S-Learner S-Learner is an approach that uses a single model with the presence or absence of intervention as a feature. Using the variable t, which represents the presence or absence of intervention, the probability of conversion (y=1) in the cases with intervention (t=1) and without intervention (t=0) is estimated, as shown in equation (2).

number

number

[0032] (A-2) T-Learner The T-Learner approach uses two separate models: one for intervention (t=1) and one for no intervention (t=0). Using these separate models, the probability of conversion is estimated as shown in equation (4).

number

number

[0033] (A-3) Class conversion Class transformation defines the target variable Z, which represents the expected result, as shown in equation (6) below.

number

number

[0034] The conventional uplift score will be explained using a numerical example. Here, we assume a case where the conversion is "purchase," and there are two user groups: the first user group with user characteristic = X1, and the second user group with user characteristic = X2. For each of the first and second user groups, the probability of purchase with intervention and the probability of purchase without intervention are as follows. First user group: Probability of purchase with intervention = 0.7 Probability of purchasing without intervention = 0.6 Second user group: Probability of purchasing with intervention = 0.3 Probability of purchase without intervention = 0.2

[0035] According to the conventional method for calculating the uplift score, the uplift score is 0.7-0.6=0.1 for the first user group and 0.3-0.2=0.1 for the second user group, meaning that the uplift score is 0.1 for both user groups. In other words, the uplift score is the same for both the first and second user groups.

[0036] On the other hand, the overall average purchase probability for the first and second user groups is ((0.7+0.6) / 2)=0.65 for the first user group and ((0.3+0.2) / 2)=0.25 for the second user group, with the first user group having a higher probability. This means that the first user group is more likely to make a purchase regardless of whether or not there is an intervention. What can be seen from these numerical results is that even if the uplift score, which can be calculated from the difference between the probability of making a purchase with and without intervention, is the same, there can be a difference in the average purchase probability. Since a higher average purchase probability indicates that the user group as a whole is more likely to make a purchase regardless of whether or not there is an intervention, it can be said that in the above example, it is preferable to select the first user group as the target.

[0037] Thus, while conventional uplift score calculation methods use the difference between the probability of purchase with and without intervention, the calculated uplift score does not reflect the purchase probability of the entire user group, regardless of whether they received intervention or not. Therefore, there was a possibility that the target user group could not be appropriately selected.

[0038] (B) Method for calculating uplift according to this embodiment This section describes an uplift calculation method according to this embodiment, which is configured to solve the above problems. In this embodiment, we propose a method for calculating an uplift score using the probability that the entire user group achieved conversion. For the purpose of explanation, below, the intervention user group, which is the group of users with intervention (t=1), will be represented as t, and the control user group, which is the group of users without intervention (t=0), will be represented as c.

[0039] First, for a group of users with user characteristics = X, the probability that a conversion occurs (y=1) among the intervention user group t, p(y=1|t,X), can be expressed as shown in equation (8) based on Bayes' theorem.

number

[0040] Next, for a group of users with user characteristics = X, the probability that a conversion occurs (y=1) among the control group of users c, p(y=1|c,X), can be expressed as shown in equation (9) based on Bayes' theorem.

number

[0041] Next, the uplift score is calculated using equation (10) with probabilities p(t|y=1,X), p(c|y=1,X), p(y=1|X), p(t|X), and p(c|X).

number

[0042] The probabilities p(y=1|t(=t=1),X) and p(y=1|c(=t=0),X) can be estimated using the S-Learner and T-Learner approaches as described above. Furthermore, the probability p(y=1|X) can be estimated based on the purchase history of the user group whose user characteristic = X. In this disclosure, the probabilities p(t|X) and p(c|X) are assumed to be constants that add up to 1. That is, the user group whose user characteristic = X is pre-divided into either an intervention-enabled or non-intervention-enabled group. For example, if the probabilities p(t|X) and p(c|X) are both 0.5, it means that for multiple users in the target population, the user group with intervention and the user group without intervention are roughly equal (or nearly equal). Therefore, the probabilities p(t|y=1,X) and p(c|y=1,X) can be estimated from equations (8) and (9).

[0043] As can be understood from equation (10), the uplift score according to this disclosure is calculated using the purchase probability p(y=1|X) for all user groups with user characteristic = X, that is, all user groups with and without intervention that have user characteristic = X. Next, we will explain the two methods for calculating the uplift score expressed by equation (10).

[0044] (B-1) Retrospective method This section describes a method for calculating the uplift score represented by equation (10) using the retrospective method described in Non-Patent Document 1. This method uses two ML models, the first ML model and the second ML model.

[0045] The first ML model is an ML model trained to estimate probabilities p(t|y=1,X) and p(c|y=1,X) according to the retrospective uplift model described in Non-Patent Document 1. Probabilities p(t|y=1,X) and p(c|y=1,X) are the probabilities of whether a user with user characteristic X who has converted belongs to the intervention user group t or the control user group c, respectively. The information processing device 10 can train the first ML model using only data with conversion (y=1). This reduces the amount of data used for estimation. Probabilities p(t|y=1,X) and p(c|y=1,X) are estimated using the trained first ML model. The second ML model is one that has been trained to estimate the probability p(y=1|X). The probability p(y=1|X) is estimated using the trained second ML model. The information processing device 10 calculates the uplift score using equation (10) with the probabilities p(t|y=1,X), p(c|y=1,X), and p(y=1|X) estimated as described above, along with the constant probabilities p(t|X) and p(c|X).

[0046] (B-2) Direct classification method In the direct classification method, a single ML model is trained using the following three class labels. The three class labels are defined as follows:

number

[0047] Using these three class labels, the information processing device 10 trains an ML model to estimate the following three probabilities.

number

[0048] The uplift score using the direct classification method will be explained with a numerical example. We will consider a case where the conversion is "purchase" and there are two user groups: the first user group with user attribute = X1 and the second user group with user attribute = X2. For the first and second user groups, the probabilities of p(0), p(1), and p(2) are estimated as follows. First user group: p(0)=0.5, p(1)=0.4, p(2)=0.1 Second user group: p(0)=0.2, p(1)=0.1, p(2)=0.7

[0049] When probabilities p(t|X) and p(c|X) are both 0.5, the uplift score calculated by equation (10) is 0.045 for the first user group and 0.015 for the second user group. According to the uplift score calculation method of this embodiment, it is possible to calculate an uplift score that reflects the purchase probability for both the first user group and the second user group.

[0050] Although two methods have been described above, the information processing device 10 may estimate the probabilities p(t|y=1,X), p(c|y=1,X), and p(y=1|X) by other methods. The information processing device 10 may then calculate the uplift score using equation (10) with the estimated probabilities p(t|y=1,X), p(c|y=1,X), and p(y=1|X), along with the constant probabilities p(t|X) and p(c|X).

[0051] [Conceptual diagram of the uplift score calculation process] Figure 2 shows a conceptual diagram of the uplift score calculation process according to this embodiment. As described above, the uplift score is calculated as shown in equation (10). Here, as shown in Figure 2, estimated values ​​are used for probabilities p(t|y=1,X), p(c|y=1,X), and p(y=1|X), while predetermined constants are used for probabilities p(t|X) and p(c|X). Then, the uplift score is calculated using equation (10) with these values.

[0052] [Configuration of the e-commerce server] Figure 3 shows an example of the functional configuration of the e-commerce server 11 according to this embodiment. Here, only the functional configuration related to the processing of this embodiment is shown, and other functional configurations, including those for providing e-commerce services, are omitted. As an example of its functional configuration, the e-commerce server 11 has a user attribute acquisition unit 301, a conversion information acquisition unit 302, a user information provision unit 303, an intervention user determination unit 304, and an advertising distribution unit 305. For the purpose of explanation, a user device 12 and user 14 are referred to, but the same explanation can be applied to other user devices and users that can communicate with the e-commerce server 11.

[0053] The user attribute acquisition unit 301 acquires one or more user attributes of user 14 associated with user ID from user device 12. The conversion information acquisition unit 302 acquires conversion information of user 14 in the e-commerce mall provided by e-commerce server 11, associated with user ID from user device 12 (for example, conversion history with time information and whether or not there have been conversions over a certain period). The user information provision unit 303 provides one or more user attributes of user 14 associated with user ID and the conversion information to the information processing device 10. The user information provision unit 303 may provide user attributes and conversion information of user 14 to the information processing device 10 each time they are acquired, or it may provide user attributes and conversion information of user 14 that have been accumulated over a certain period to the information processing device 10 each time they are acquired. The intervention user determination unit 304 obtains the uplift score for each user group calculated by the information processing device 10 and determines which users will be intervention users, i.e., the intervention user group (also referred to as target users). Specifically, the intervention user determination unit 304 determines the intervention user group, which will be the intervention user group. The ad delivery unit 305 functions as the intervention unit that executes the intervention and delivers advertisements to the intervention user group that encourage conversions such as purchases. Specifically, the ad delivery unit 305 delivers advertisements to user IDs identified in the intervention user group. The advertisements may be generated by the ad delivery unit 305, generated by a predetermined function in the e-commerce server 11, or generated and obtained by other devices. In this embodiment, the ad delivery unit 305 executes the intervention by delivering advertisements, but it may be configured to execute other forms of intervention.

[0054] [Configuration of the information processing device] Figure 4 shows an example of the functional configuration of the information processing device 10 according to this embodiment. As an example of its functional configuration, the information processing device 10 includes a user information acquisition unit 401, a user group generation unit 402, an uplift score calculation unit 403, an uplift score provision unit 404, a user information storage unit 410, and a learning model storage unit 420. The learning model storage unit 420 is configured to store one or more machine learning models (for example, data representing architecture and various parameters) used to calculate the uplift score. Furthermore, although we will refer to user device 12 and user 14 for the purpose of explanation, the same explanation can be applied to other user devices and users.

[0055] The user information acquisition unit 401 acquires user information from the e-commerce server 11, including one or more user attributes of user 14 associated with the user ID of user 14, and conversion information. The user information acquisition unit 401 stores the acquired user information in the user information storage unit 410. The user group generation unit 402 generates one or more user groups by grouping one or more users having one or more common (identical) user attributes based on the user attributes stored in the user information storage unit 410. The uplift score calculation unit 403 calculates an uplift score for each user group generated by the user group generation unit 402. The method for calculating the uplift score is as described above. The uplift score calculation unit 403 can calculate the uplift score using one or more machine learning models stored in the learning model storage unit 420. The uplift score provision unit 404 provides the calculated uplift score for each user group to the e-commerce server 11.

[0056] [Hardware configuration of information processing equipment] Next, an example of the hardware configuration of the information processing device 10 will be described. Figure 5 is a block diagram showing an example of the hardware configuration of the information processing device 10 according to this embodiment. The hardware configuration of the e-commerce server 11 is similar. The information processing device 10 according to this embodiment can be implemented on one or more computers, mobile devices, or any other processing platform. Referring to Figure 5, an example is shown in which the information processing device 10 is implemented in a single computer; however, the information processing device 10 according to this 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.

[0057] As shown in Figure 5, the information processing device 10 may include a CPU (Central Processing Unit) 501, a ROM (Read Only Memory) 502, a RAM (Random Access Memory) 503, an HDD (Hard Disk Drive) 504, an input unit 505, a display unit 506, a communication interface 507, and a system bus 508. The information processing device 10 may also include external memory. The CPU 501 comprehensively controls the operation of the information processing device 10 and controls each component (502-507) via the system bus 508, which is a data transmission path.

[0058] ROM 502 is a non-volatile memory that stores control programs and other information necessary for the CPU 501 to execute processing. This program includes instructions (code) that cause the processing according to the above embodiment to be executed. This program may be stored in non-volatile memory such as HDD 504 or SSD (Solid State Drive), or in external memory such as a removable storage medium (not shown). RAM 503 is volatile memory and functions as the main memory, work area, etc., of the CPU 501. In other words, when executing processing, the CPU 501 loads necessary programs, etc., from ROM 502 into RAM 503 and executes these programs, etc., to realize various functional operations. RAM 503 may include the user information storage unit 410 and the learning model storage unit 420 shown in Figure 4.

[0059] HDD504 stores various data and information necessary for CPU501 to perform processing using programs, for example. Furthermore, HDD504 also stores various data and information obtained through processing performed by CPU501 using programs, for example. The input unit 505 consists of a pointing device such as a keyboard or mouse. The display unit 506 is comprised of a monitor such as a liquid crystal display (LCD). The display unit 506 may also function as a GUI (Graphical User Interface) when configured in combination with the input unit 505.

[0060] Communication I / F 507 is an interface that controls communication between the information processing device 10 and an external device. Communication I / F 507 provides an interface to a network and performs communication with the external device via the network. Various data and parameters are sent and received between the external device and the communication I / F 507. In this embodiment, communication I / F 507 may perform communication via a wired LAN (Local Area Network) or a dedicated line compliant 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 consist of 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). Furthermore, it includes wireless WANs (Wide Area Networks) such as 4G and 5G. Furthermore, the network only needs to connect each device in a way that allows it to communicate with one another, and the communication standards, scale, and configuration are not limited to those described above.

[0061] At least some of the functions of the information processing device 10 shown in Figure 4 can be realized by the CPU 501 executing a program. However, at least some of the functions of the information processing device 10 shown in Figure 4 may be operated as dedicated hardware. In this case, the dedicated hardware operates based on the control of the CPU 501.

[0062] [Processing flow in information processing systems] Figure 6 shows a flowchart of the processes performed in the information processing system according to this embodiment. It is assumed that processes S61, S62, S66, and S67 are performed by the e-commerce server 11, and processes S63 to S65 are performed by the information processing device 10. However, if the information processing device 10 and the e-commerce server 11 are configured as a single device, they may be performed by that device.

[0063] In S61, the user attribute acquisition unit 301 of the e-commerce server 11 acquires one or more user attributes of multiple users at any time, and the conversion information acquisition unit 302 acquires conversion information of multiple users at any time. In S62, the user information provision unit 303 of the e-commerce server 11 provides the user attributes and conversion information to the information processing device 10 as user information. The user information is acquired by the user information acquisition unit 401 of the information processing device 10 and stored in the user information storage unit 410.

[0064] In S63, the user group generation unit 402 of the information processing device 10 generates one or more user groups by grouping one or more users having one or more common (identical) user attributes based on the user attributes stored in the user information storage unit 410. For the purpose of explanation, it is assumed that the user group generation unit 402 generates a user group where user characteristic = X, and for the purpose of explaining Figure 6, this user group will also be referred to as user group X. The user group generation unit 402 groups user group X into a first user group t (i.e., the intervening user group) that has undergone intervention to induce a predetermined conversion, and a second user group c (i.e., the control user group) that has not undergone the intervention. In this embodiment, the grouping is done so that the first user group t and the second user group c each account for half (or approximately half). As a result, the probabilities p(t|X) and p(c|X) are each 0.5.

[0065] In S64, the uplift score calculation unit 403 calculates an uplift score for the user group X generated by the user group generation unit 402. In this embodiment, as described above, the uplift score calculation unit 403 estimates the probability p(t|y=1,X) that a user who converted is included in the first user group t, and the probability p(c|y=1,X) that a user who converted is included in the second user group c, using means such as a machine learning model. The uplift score calculation unit 403 also estimates the probability p(y=1|X) that the entire user group X converted, using means such as a machine learning model, as described above. For explanation purposes, the probabilities p(t|y=1,X), p(c|y=1,X), and p(y=1|X) are referred to as the first probability, second probability, and third probability, respectively. Subsequently, the uplift score calculation unit 403 calculates an uplift score based on the first probability, second probability, and third probability. Specifically, the uplift score calculation unit 403 uses the first probability, the second probability, and the third probability, along with probabilities p(t|X) and p(c|X), to calculate an uplift score for user group X according to equation (10). The information processing device 10 calculates an uplift score for one or more user groups through processing from S63 to S65, and provides the calculated uplift score for each user group to the e-commerce server 11. The provided uplift scores for one or more user groups are acquired by the intervention user determination unit 304 of the e-commerce server 11.

[0066] In S66, the intervention user determination unit 304 of the e-commerce server 11 determines the intervention user group based on the uplift score for one or more user groups. For example, the intervention user determination unit determines the intervention user group to be a group of users whose uplift score is higher than a predetermined value. The predetermined value is, for example, 0.5. Subsequently, in S67, the advertising distribution unit 305 of the e-commerce server 11 performs an intervention on the intervention user group. For example, the advertising distribution unit 305 distributes advertisements to the intervention user group that encourage conversions such as purchases. The advertising distribution unit 305 may also perform an intervention using the user characteristics of the intervention user group. For example, the advertising distribution unit 305 may distribute advertisements that encourage conversions of products related to the user characteristics of the intervention user group. The advertising distribution unit 305 may also distribute advertisements that encourage conversions of products related to similar user characteristics that are similar to the user characteristics of the intervention user group. These similar user characteristics can be obtained by machine learning.

[0067] Thus, since the uplift score is calculated by considering the probability that all users made a purchase, the calculated uplift score can serve as an indicator of purchase trends. Furthermore, because the calculated uplift score represents purchase trends, interventions using the uplift score can be expected to improve ROI (Return on Investment).

[0068] Although specific embodiments are described above, these embodiments are merely illustrative and not intended to limit the scope of the present invention. Apparatuses and methods described herein can be embodied in forms other than those described above. Furthermore, the embodiments described above can be appropriately omitted, substituted, and modified without departing from the scope of the present invention. Such omitted, substituted, and modified forms fall within the scope of the claims and their equivalents and are within the technical scope of the present invention.

[0069] This embodiment includes the following configuration. [1] The system includes a grouping unit that groups a user group consisting of multiple users into a first user group that has undergone an intervention to induce a predetermined conversion and a second user group that has not undergone the intervention; a first estimation unit that estimates a first probability that the first user group achieved the conversion and a second probability that the second user group achieved the conversion; a second estimation unit that estimates a third probability that the entire user group achieved the conversion; and a calculation unit that uses the first, second, and third probabilities to calculate an uplift score that represents the effect of the intervention on the user group.

[0070] [2] The information processing apparatus according to [1], wherein the user group has one or more common user attributes.

[0071] [3] The information processing apparatus according to [1] or [2], further comprising an intervention unit that performs the intervention on the user group when the uplift score is higher than a predetermined value.

[0072] [4] The information processing apparatus according to [3], wherein the user group has one or more common user attributes, and the intervention unit performs the intervention using the one or more common user attributes of the user group.

[0073] [5] The information processing device according to any one of [1] to [4], wherein the conversion is the purchase of an item offered in a service for electronic commerce.

[0074] [6] The information processing apparatus according to any one of [1] to [5], wherein the intervention is to deliver an advertisement relating to the service.

[0075] [7] The information processing apparatus according to any one of [1] to [6], wherein the first estimation unit estimates the first probability and the second probability using a first machine learning model, and the second estimation unit estimates the third probability using a second machine learning model.

[0076] [8] The information processing apparatus according to any one of [1] to [6], wherein the first estimation unit estimates the first probability and the second probability using a first machine learning model, and the second estimation unit estimates the third probability using the first machine learning model. [Explanation of symbols]

[0077] 10: Information processing device, 11: E-commerce server, 12: User device, 13: Network, 14: User, 301: User attribute acquisition unit, 302: Conversion information acquisition unit, 303: User information provision unit, 304: Intervention user determination unit, 305: Ad delivery unit, 401: User information acquisition unit, 402: User group generation unit, 403: Uplift score calculation unit, 404: Uplift score provision unit, 410: User information storage unit, 420: Learning model storage unit

Claims

1. A grouping unit that groups a user group consisting of multiple users into a first user group that has undergone intervention to stimulate a predetermined conversion, and a second user group that has not undergone the said intervention, A first estimation unit estimates a first probability, which is the probability that a user who has achieved conversion is included in the first user group, and a second probability, which is the probability that a user who has achieved conversion is included in the second user group. A second estimation unit estimates a third probability, which is the probability that the entire user group achieved the conversion. A calculation unit that uses the first probability, the second probability, and the third probability to calculate an uplift score representing the effect of the intervention on the user group, An information processing device having

2. The user group has one or more common user attributes, The information processing apparatus according to claim 1.

3. The system further includes an intervention unit that performs the intervention on the user group when the uplift score is higher than a predetermined value. The information processing apparatus according to claim 1.

4. The user group has one or more common user attributes, and the intervention unit performs the intervention using the one or more common user attributes of the user group. The information processing apparatus according to claim 3.

5. The aforementioned conversion is the purchase of an item offered through a service for e-commerce. The information processing apparatus according to claim 1.

6. The intervention described above involves delivering advertisements related to the service. The information processing apparatus according to claim 1.

7. The first estimation unit estimates the first probability and the second probability using the first machine learning model, The second estimation unit estimates the third probability using a second machine learning model. The information processing apparatus according to claim 1.

8. The first estimation unit estimates the first probability and the second probability using the first machine learning model, The second estimation unit estimates the third probability using the first machine learning model. The information processing apparatus according to claim 1.

9. An information processing method performed by an information processing device, This involves grouping a user group consisting of multiple users into a first user group that received intervention to stimulate a predetermined conversion, and a second user group that did not receive the said intervention. To estimate a first probability, which is the probability that a user who has achieved the conversion is included in the first user group, and a second probability, which is the probability that a user who has achieved the conversion is included in the second user group, To estimate a third probability, which is the probability that the entire user group reached the conversion, Using the first probability, the second probability, and the third probability, an uplift score representing the effect of the intervention on the user group is calculated, Information processing methods, including those mentioned above.

10. An information processing program for causing a computer to perform information processing, wherein the program causes the computer to perform information processing. A grouping process that divides a user group consisting of multiple users into a first user group that has undergone intervention to stimulate a predetermined conversion, and a second user group that has not undergone the said intervention, A first estimation process that estimates a first probability, which is the probability that a user who has achieved conversion is included in the first user group, and a second probability, which is the probability that a user who has achieved conversion is included in the second user group. A second estimation process that estimates a third probability, which is the probability that the entire user group reached the conversion, This involves performing a calculation process that includes calculating an uplift score representing the effect of the intervention on the user group using the first probability, the second probability, and the third probability. Information processing program.

Images