Information processing device
The information processing device uses user data to generate a learning model that predicts vehicle purchases by combining variables from both the same and different sellers, addressing the challenge of identifying potential buyers among numerous campaign applicants.
Patent Information
- Application Number
- JP2020162147
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2020-09-28
- Publication Date
- 2024-11-27
- Estimated Expiration
- 2040-09-28
AI Technical Summary
Vehicle sales companies face challenges in narrowing down a large number of campaign applicants to identify potential buyers accurately for targeted sales activities.
An information processing device that generates a learning model using user data, including browsing history and purchase information, to derive a score indicating the likelihood of purchasing a vehicle, by combining variables from both the same and different sellers.
Accurately narrows down users likely to purchase vehicles, enabling targeted sales activities with high accuracy.
Smart Images

Figure 0007590841000001 
Figure 0007590841000002 
Figure 0007590841000003
Abstract
Description
[Technical field]
[0001] The present invention relates to an information processing device. [Background technology]
[0002] Patent document 1 proposes an information processing device that performs a category determination of a URL indicating the location of a web page, calculates and learns the user's level of interest based on the category determination results, and provides information to the user's computer according to the level of interest. [Prior art documents] [Patent documents]
[0003] [Patent Document 1] JP 2008-176511 A Summary of the Invention [Problem to be solved by the invention]
[0004] For example, a vehicle sales company (dealer) or a vehicle manufacturer may conduct a campaign to provide financial assistance for purchasing a vehicle. In such a campaign, a user accesses a website provided by the sales company or the vehicle manufacturer and applies for the campaign on the website.
[0005] Users who wish to apply for a campaign answer a questionnaire on a website in order to apply for the campaign. The sales company or vehicle manufacturer selects the campaign winners from among the users who applied for the campaign. In addition, the sales company may also conduct sales activities for vehicle sales to users who are not selected for the campaign, based on the results of the questionnaire.
[0006] However, there are many users who apply for the campaign, and it is difficult to conduct sales activities for all of the users. Therefore, there is a need to narrow down the users who are likely to make a purchase.
[0007] The present invention aims to provide an information processing device capable of accurately generating a learning model for narrowing down users who are likely to make a purchase. [Means for solving the problem]
[0008] In order to solve the above problem, an information processing device of the present invention includes an acquisition unit that acquires variable data associated with a user, the variable being information included in first information consisting of information acquired from a user and second information consisting of a browsing history of a web page by the user; a learning unit that generates a learning model based on first variable data, among the variable data acquired by the acquisition unit, for which a purchase result of the user is known, and whether or not the user associated with the first variable data has purchased a first product; and a derivation unit that uses the learning model to derive a score for second variable data, among the variable data acquired by the acquisition unit, for which a purchase result of the user is not known, wherein the acquisition unit: The above The variable data is obtained by using third information, which is information regarding the number of times a web page related to a second product different from the first product, included in the second information, as a variable. and acquiring, as the third information, the variable data having a variable being the number of views of the category into which the second product is classified, and acquiring, for a third product included in the first information or the second information, the variable data having the third product as a variable without classifying the third product into the category, the second product being a product sold by a seller different from that of the first product, and the third product being a product sold by the same seller as that of the first product. . In order to solve the above problem, the information processing device of the present invention includes an acquisition unit that acquires variable data associated with the user, the variable data being composed of information acquired from the user, the first information being different from information regarding the user's webpage browsing history and different from information regarding the user's purchasing results, and information included in second information consisting of the user's webpage browsing history; a learning unit that generates a learning model based on the first information and the second information included in the first variable data acquired by the acquisition unit, for which the user's purchasing results are known, and whether or not the user associated with the first variable data has purchased a first product; and a derivation unit that uses the learning model to derive a score for second variable data, for which the user's purchasing results are not known, among the variable data acquired by the acquisition unit, and the acquisition unit acquires the variable data having third information, the third information being information regarding the number of times a webpage related to a second product different from the first product, included in the second information, as a variable.
[0010] The derivation unit may derive a purchase probability of a fourth product as the score, and the fourth product may be a product sold by the same seller as the first product.
[0011] The first product may be a product sold by the company itself, and the second product may be a product sold by another company.
[0012] The acquisition unit may combine a plurality of variables included in the variable data and acquire them as one variable. Effect of the Invention
[0013] According to the present invention, it is possible to generate a learning model with high accuracy that narrows down users who are likely to make a purchase. [Brief description of the drawings]
[0014] [Figure 1] FIG. 1 is an explanatory diagram showing a schematic configuration of an information processing system. [Diagram 2] FIG. 2 is an explanatory diagram showing a functional configuration of the information processing system. [Diagram 3] FIG. 2 is a diagram illustrating an example of a web page provided by a web server. [Figure 4] FIG. 2 is a diagram illustrating an example of a web page provided by a web server. [Diagram 5] FIG. 2 is a diagram illustrating an example of a web page provided by a web server. [Figure 6] FIG. 13 is a diagram illustrating variable data. [Figure 7] FIG. 13 is a diagram illustrating a competitor vehicle model name category table. [Figure 8] FIG. 13 is a diagram illustrating teacher data. [Figure 9] FIG. 11 is a first diagram illustrating variable data in another embodiment. [Figure 10] FIG. 2 is a second diagram illustrating variable data in another embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0015] Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. The dimensions, materials, and other specific numerical values shown in the embodiment are merely examples for facilitating understanding of the invention, and do not limit the present invention unless otherwise specified. In this specification and drawings, elements having substantially the same functions and configurations are given the same reference numerals to avoid repeated explanations, and elements not directly related to the present invention are not shown.
[0016] 1 is an explanatory diagram showing a schematic configuration of an information processing system S. The information processing system S includes an information processing device 1, a web server 2, and a user terminal 3.
[0017] The information processing device 1, the web server 2, and the user terminal 3 can establish communication via a communication network 4. Note that the information processing device 1 only needs to be able to establish communication with at least the web server 2, and does not need to be able to establish communication with the user terminal 3. Also, the user terminal 3 only needs to be able to establish communication with at least the web server 2, and does not need to be able to establish communication with the information processing device 1.
[0018] The information processing device 1 and the web server 2 are managed by the same or related company. For example, the information processing device 1 and the web server 2 are managed by a vehicle manufacturer (here, "Company A" that manufactures vehicles). Note that, although the following description will be given of a case where Company A's vehicles are sold by a dealer (seller), Company A may also be the seller and sell the vehicles.
[0019] The web server 2 provides web pages with information on vehicles (products) manufactured by the company A, campaigns to subsidize the purchase of vehicles of the company A, and the like.
[0020] The user terminal 3 broadly includes electronic devices capable of wireless or wired communication connection at least with the web server 2. Examples of the user terminal 3 include a smartphone, a mobile phone, a tablet device, and a personal computer.
[0021] The communication network 4 is composed of a mobile phone network, the Internet, a LAN (Local Area Network), a dedicated line, etc., and enables communication between the information processing device 1, the web server 2, and the user terminal 3.
[0022] FIG. 2 is an explanatory diagram showing the functional configuration of the information processing system S.
[0023] The information processing device 1 is a computer including a CPU (Central Processing Unit) which is an arithmetic processing device, a ROM (Read Only Memory) which is a storage element for storing programs and arithmetic parameters used by the CPU, and a RAM (Random Access Memory) which is a storage element for temporarily storing parameters that change appropriately during execution of the CPU. As shown in Fig. 2, the information processing device 1 functions as a data acquisition unit 11, a learning unit 12, a derivation unit 13, a data storage unit 14, and a communication unit 15. Each unit of the information processing device 1 will be described in detail later.
[0024] The web server 2 is a computer including a CPU, a ROM, a RAM, etc., and functions as a control unit 21, a communication unit 22, and a data storage unit .
[0025] The control unit 21 performs overall control of the web server 2. The communication unit 22 establishes communication with the information processing device 1 and the user terminal 3 via the communication network 4. The data storage unit 23 stores data of web pages provided by the web server 2.
[0026] The user terminal 3 is a computer including a CPU, a ROM, a RAM, etc., and functions as a control unit 31, a communication unit 32, and a data storage unit 33.
[0027] The control unit 31 controls the entire user terminal 3. The communication unit 32 establishes communication with the web server 2 via the communication network 4. The data storage unit 33 stores browser software for browsing web pages and the like.
[0028] The user terminal 3 also includes an input unit 34 such as a keyboard or a touch panel, and a display unit 35 such as a liquid crystal display or an organic EL display.
[0029] The user terminal 3 acquires the data of a web page stored in the data storage unit 23 of the web server 2 by specifying a URL in the browser software, and displays on the display unit 35 a web page based on the acquired data.
[0030] 3 to 5 are diagrams illustrating examples of web pages provided by the web server 2. As described above, the data storage unit 23 of the web server 2 stores web page data such as information on vehicles manufactured by company A and campaigns to provide financial assistance when purchasing a vehicle from company A.
[0031] In the user terminal 3, by specifying the URL of a web page provided by the web server 2, the web page shown in Fig. 3 to Fig. 5 is displayed on the display unit 35. For example, as shown in Fig. 3, the data storage unit 23 of the web server 2 stores data of a user registration page 40 for user registration to receive various services provided by company A.
[0032] When the user terminal 3 acquires the data of the user registration page 40, the browser software displays the user registration page 40 shown in Fig. 3 on the display unit 35. The user registration page 40 has input fields for inputting answers to items such as name, address, date of birth, occupation, name of the owned vehicle (name of the model owned), and name of the dealer. Note that the items provided on the user registration page 40 are merely examples, and some of these may be omitted, or the page may further include vehicle information such as the acquisition date of the owned vehicle, and personal information such as an email address.
[0033] Then, when an answer to each item is entered via input unit 34 into an input field provided on user registration page 40 and the registration button is operated, the information entered into the input field (hereinafter referred to as registration information) is transmitted to web server 2 via communication unit 32. Web server 2 stores the transmitted registration information in data storage unit 23 and generates a unique ID (Company A ID) for the user who transmitted the registration information. The generated ID is transmitted to user terminal 3 via communication unit 22.
[0034] As a result, the user of the user terminal 3 can obtain the transmitted ID and use the obtained ID to receive various subsequent services.
[0035] As shown in FIG. 4, the data storage unit 23 of the web server 2 also stores data for a campaign page (hereinafter, called campaign page 41) that provides financial assistance when purchasing a vehicle from company A.
[0036] When the user terminal 3 acquires the data of the campaign page 41, the browser software displays the campaign page 41 shown in Fig. 4 on the display unit 35. The campaign page 41 has input fields for inputting answers to items such as ID, name, address, date of birth, occupation, name of the car model owned, whether or not you are considering replacing your car, whether or not you have a dealership (dealer) with whom you have a relationship, the name of the dealership, whether or not you are interested in a car from Company A, the name of the car model you are interested in, your ownership history of Company A's cars, the name of the car model owned by Company A, whether or not you will consider a car from Company A for your next purchase, when you will next purchase, the main user, whether or not you have been to Company A's dealership, the period until the next vehicle inspection, and the number of years since the owner's ownership. Note that the items provided on the campaign page 41 are merely examples, and some of these items may be omitted or other items may be provided.
[0037] Then, when an answer to each item is entered into an input field provided on campaign page 41 via input unit 34 and the registration button is operated, the information entered into the input field (hereinafter referred to as campaign information) is transmitted to web server 2 via communication unit 32. Web server 2 stores the transmitted campaign information in data storage unit 23.
[0038] When the campaign period ends, winners are determined based on the campaign information stored in the data storage unit 23, and the determined winners (users) are notified of their win. The winners may be determined by lottery by the web server 2, by the information processing device 1, or by other methods.
[0039] The data storage unit 23 of the web server 2 also stores data for a vehicle information page that displays information on the vehicles (models) of Company A. This vehicle information page includes a competitor's vehicle comparison page 42 that compares the specifications, etc., of Company A's vehicles with those of other companies' vehicles (vehicles sold by other companies' dealers).
[0040] In the user terminal 3, when the model name of the vehicle of company A is selected and the model name of the vehicle of another company to be compared (here, car a of company a) is selected, the browser software displays a other company vehicle comparison page 42 shown in FIG. 5 on the display unit 35. The other company vehicle comparison page 42 displays a list of the model name of the vehicle of company A selected by the user and the specifications of the model name of the vehicle of the other company. For example, the vehicle manufacturer, model name, body shape, number of doors, drive system, transmission, sales start date, vehicle price, overall length, overall width, overall height, etc. are displayed in a list on the other company vehicle comparison page 42. Note that the specifications displayed in a list on the other company vehicle comparison page 42 are merely an example, and some of these may be omitted or other information may be included. In addition, in the example shown in FIG. 5, there is one vehicle of company A and one vehicle of another company, but there may be multiple vehicles of company A and other companies.
[0041] Meanwhile, in the user terminal 3, a cookie is generated and updated by the browser software every time a web page is viewed. The cookie includes an identifier for identifying the user terminal 3 and browsing history information (second information) of the viewed web page.
[0042] Then, every time the user terminal 3 browses a web page provided by the web server 2, it transmits the cookie of that web page to the web server 2. In this embodiment, the user terminal 3 updates the cookie and transmits it to the web server 2 when at least the user registration page 40, the campaign page 41, and the other company's vehicle comparison page 42 are browsed.
[0043] When a user registers on the user registration page 40, the web server 2 stores in the data storage unit 23 an identifier for identifying the user terminal 3 that performed the user registration and the registration information in association with each other based on the cookie.
[0044] In addition, when the web server 2 acquires campaign information via the campaign page 41, the web server 2 associates the campaign information with an identifier for identifying the user terminal 3 that acquired the campaign information based on the cookie, and stores the associated information in the data storage unit 23.
[0045] In addition, when the web server 2 acquires browsing history information via the other company vehicle comparison page 42, the web server 2 associates the browsing history information with an identifier for identifying the user terminal 3 that acquired the browsing history information based on the cookie, and stores the associated information in the data storage unit 23.
[0046] Therefore, in the web server 2, the registration information, campaign information, and browsing history information that have the same identifier for identifying the user terminal 3 are stored in the data storage unit 23 in an associated manner.
[0047] Fig. 6 is a diagram for explaining variable data. Note that the variables included in the variable data shown in Fig. 6 are only a part of the data. The data acquisition unit 11 of the information processing device 1 acquires the registration information, campaign information, and browsing history information stored in the data storage unit 23 of the web server 2 via the communication unit 15. Note that when acquiring the registration information, campaign information, and browsing history information, the data acquisition unit 11 also acquires an identifier for identifying the user terminal 3, which is associated with the information.
[0048] Then, the data acquisition unit 11 generates variable data for each user based on the acquired registration information, campaign information, browsing history information, and an identifier for identifying the user terminal 3.
[0049] Specifically, as shown in Fig. 6, the data acquisition unit 11 regards each piece of information included in the registration information, campaign information, and browsing history information that have the same identifier for identifying the user terminal 3 as a variable, and collects the variables to generate variable data for one user and store it in the data storage unit 14. Note that if the registration information and the campaign information contain the same information (e.g., occupation), the same information is collected into one. Age information is derived from the date of birth included in the registration information.
[0050] In this way, the data acquiring unit 11 generates variable data by associating each piece of information acquired from the same user terminal 3 and consolidating it as a variable for the same user.
[0051] 7 is a diagram for explaining the other company's vehicle name category table. As described above, on the other company's vehicle comparison page, by selecting the other company's vehicle (model name) that the user wants to compare with the vehicle (model name) of company A, the specifications of the vehicle of company A selected by the user and the other company's vehicle are displayed in a list. The browsing history information includes the selected other company's vehicle (model name).
[0052] Here, there are a huge number of vehicles (model names) of other companies. Therefore, the data acquisition unit 11 classifies the vehicles of other companies into categories and generates variable data with the number of views for each category as a variable. Specifically, the data acquisition unit 11 refers to the other company's model name category table shown in FIG. 7, derives the number of views for each category to which the vehicles of other companies (model names) belong as a variable based on the viewing history information, and stores it in the variable data. For example, cars a, b, and c are classified as compact sedans, cars d, e, and f are classified as SUVs, cars g and h are classified as minivans, and cars i and j are classified as 1BOX. Then, for example, in the other company's vehicle comparison page, if car a is selected (viewed) three times, car b is selected (viewed) once, and car c is selected (viewed) once, five times is stored in the variable data as a variable for compact sedans into which cars a, b, and c are classified.
[0053] In this way, by classifying other companies' vehicles (second products) into categories and treating the number of times other companies' vehicles in each category (third information) are viewed as a variable, the degree of user interest in each category can be derived as a variable. Also, since the number of variables is reduced compared to treating other companies' vehicles as separate variables, it becomes easier to grasp the trends for each variable, and the purchase probability can be derived with high accuracy.
[0054] Fig. 8 is a diagram for explaining teacher data. The learning unit 12 extracts variable data of users who have already purchased a vehicle (first product) of company A and users who have not purchased a vehicle of company A from the variable data generated by the data acquisition unit 11, and generates teacher data (first variable data) in which the variable data is associated with a purchase result (purchased=1, not purchased=0) as shown in Fig. 8, that is, the purchase result is known.
[0055] In this training data, for example, users who have acquired campaign information in past campaigns are selected as targets. Note that the purchase results are input by company A, which manages the information processing device 1.
[0056] Then, the learning unit 12 generates a learning model by performing, for example, a logistic regression analysis using the teacher data.
[0057] In logistic regression analysis, each piece of information in the training data is a variable xi, the purchase result is a response variable y, and the partial regression variable ai is derived using equation (1) for nonlinear regression analysis, for example, by the least squares method. y=a1x1+a2x2+a3x3+···+anxn+b (1)
[0058] Then, the learning unit 12 stores the formula (1) from which the partial regression variable ai is derived as a learning model in the data storage unit 14. Note that the method for deriving the learning model is not limited to logistic regression analysis, and other methods may be used.
[0059] When the registration information, campaign information, and browsing history information newly acquired from the web server 2 are stored in the data storage unit 14 as variable data by the data acquisition unit 11, the derivation unit 13 uses a learning model to derive a target variable y for the variable data (second variable data for which the purchase or non-purchase status is not known).
[0060] Here, since the partial regression variable ai is input to the learning model, the derivation unit 13 derives the objective variable y (score) by substituting each piece of information of the variable data into the variable xi. The objective variable y takes a value between 0 and 1, and the higher this value, the higher the degree of match with the variable data of a user who has purchased a vehicle of company A, i.e., the higher the probability of purchasing a vehicle of company A. In other words, the derived objective variable y can be said to be the purchase probability of purchasing a vehicle (fourth product) of company A.
[0061] Therefore, by Company A reporting users with a high value of the objective variable y to the dealer, the dealer can focus its sales activities on users who are likely to purchase Company A's vehicles.
[0062] In this way, when acquiring variable data, the information processing device 1 acquires the number of views for each category for vehicles of other companies as a variable. On the other hand, the information processing device 1 acquires information about the vehicle of company A (third product) as a variable without classifying it into a category. Then, the information processing device 1 generates a learning model based on the acquired variable data and the purchase results. Furthermore, the information processing device 1 uses the learning model to derive a target variable y (score) for the acquired variable data.
[0063] As a result, the information processing device 1 can use the number of times of viewing classified by category as a variable for the viewing history information of a huge number of vehicles of other companies, and can use the degree of interest in the category that the user is interested in as a variable. Thus, the information processing device 1 can reflect the user's interest tendency in the learning model compared to the case where the vehicles of other companies are used as individual variables, and can accurately generate a learning model that narrows down users who are likely to purchase.
[0064] Although the present invention has been described above with reference to the accompanying drawings, it goes without saying that the present invention is not limited to such embodiments. It is clear that a person skilled in the art can think of various modified or altered examples within the scope of the claims, and it is understood that such examples also naturally belong to the technical scope of the present invention.
[0065] The registration information, campaign information, and browsing history information in the above embodiment are merely examples. In the above embodiment, the registration information and campaign information are input by the user on a web page, that is, the first information acquired from the user. The information processing device 1 only needs to acquire variable data for each user using one or both of the input information and the browsing history information as variables.
[0066] In the above embodiment, each piece of information included in the registration information, campaign information, and browsing history information is treated as one variable. However, a plurality of pieces of information may be combined and treated as one variable. For example, as shown in FIG. 9, it is possible to combine gender, age, and owned car model name into one variable. Specifically, for a user with a company A ID of 0000001, variable data is generated with a combination of gender = male, age = 50 years old, and owned car model name = car A as one variable. In such a case, the degree of match with the teacher data having a combination of gender = male, age = 50 years old, and owned car model name = car A as one variable increases, and it becomes possible to derive the objective variable y with high accuracy. As a result, when the input information and browsing history information are limited, the objective variable y can be derived with high accuracy. In addition, it is possible to increase the number of variables and improve accuracy without increasing the number of items (number of inputs) input by the user, that is, without causing the user any trouble.
[0067] In the above embodiment, the information processing device 1 acquires the registration information, campaign information, and browsing history information from the web server 2. However, these pieces of input information may be acquired by other means with the consent of the user.
[0068] In the above embodiment, the registration information and the campaign information are acquired as the input information. However, the input information may be, for example, information written on paper. In other words, the input information may be acquired from the user.
[0069] In the above embodiment, the number of views for each category is derived as information regarding the number of views of other companies' products. However, the information regarding the number of views of other companies' products is not limited to this, and may be, for example, the number of views for each other company's product (vehicle). In this case, for example, as shown in Fig. 10, variable data is generated with the number of views of car a, car b, car c, ... as variables.
[0070] In the above embodiment, the information processing device 1 and the web server 2 are configured with different hardware (computers). However, the information processing device 1 and the web server 2 may be configured with the same hardware.
[0071] In addition, in the above embodiment, a vehicle is taken as an example of a product, but the product may be something other than a vehicle. [Explanation of symbols]
[0072] 1. Information processing device 11 Data Acquisition Section 12 Learning Department 13 Derivation part
Claims
1. an acquisition unit that acquires variable data associated with a user, the variable being information included in first information consisting of information acquired from the user and second information consisting of a browsing history of web pages by the user; a learning unit that generates a learning model based on first variable data, which is a purchase result of the user, among the variable data acquired by the acquisition unit, and on whether or not the user associated with the first variable data has purchased a first product; a derivation unit that uses the learning model to derive a score for second variable data, the second variable data being a purchase result of the user that is not known, among the variable data acquired by the acquisition unit; Equipped with The acquisition unit is acquiring the variable data, the variable data being set to third information, the third information being information regarding the number of times a web page related to a second product different from the first product, which is included in the second information; As the third information, variable data is acquired, the variable being the number of times the category into which the second product is classified is viewed; Regarding a third product included in the first information or the second information, without classifying the third product into the category, the variable data is acquired, the variable data having the third product as a variable; The second product is a product sold by a different seller than the first product, An information processing device, wherein the third product is a product sold by the same seller as the first product.
2. an acquisition unit that acquires variable data associated with the user, the variable including first information, the first information being different from information regarding a browsing history of web pages by the user and different from information regarding a purchase result of the user, and information included in second information being the browsing history of web pages by the user; a learning unit that generates a learning model based on the first information and the second information included in first variable data, among the variable data acquired by the acquisition unit, in which the purchase result of the user is known, and based on whether or not the user associated with the first variable data has purchased a first product; a derivation unit that uses the learning model to derive a score for second variable data, the second variable data being a purchase result of the user that is not known, among the variable data acquired by the acquisition unit; Equipped with The acquisition unit is an information processing device that acquires the variable data, which has third information as a variable, the third information being information regarding the number of views of a web page related to a second product different from the first product, which is included in the second information.
3. The derivation unit derives a purchase probability of a fourth product as the score, and The information processing device according to claim 1 , wherein the fourth product is sold by the same seller as the first product.
4. The information processing device according to claim 1 , wherein the first product is a product sold by the company, and the second product is a product sold by another company.
5. The information processing device according to claim 1 , wherein the acquisition unit combines a plurality of variables included in the variable data and acquires the combined variables as one variable.
Citation Information
Patent Citations
Commodity sales quantity prediction system
JP2000200260A
Internet advertising method and advertising system
JP2007079759A
Information processing method in computer network and information processor
JP2008176511A
Methods and systems for selecting, filtering, and / or presenting available retailers.
JP2014521147A
Learning device, learning method, learning program, learning data, and model
JP2019020930A
