DRIVER TENDENCY PREDICTION DEVICE, DRIVER TENDENCY PREDICTION METHOD, LEARNING DEVICE, LEARNING METHOD, AND INFORMATION PROCESSING PROGRAM

The system predicts a driver's tendency using user and vehicle characteristics from multiple services, enhancing prediction accuracy and enabling tailored insurance services.

JP7583202B2Active Publication Date: 2024-11-13RAKUTEN GROUP INC
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
JP2024007152
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2024-01-22
Publication Date
2024-11-13
Estimated Expiration
2042-12-19

AI Technical Summary

Technical Problem

Existing methods for determining a driver's tendency, such as good or bad driving habits, rely solely on objective data from vehicle usage and accident history, lacking a mechanism to predict tendencies based on the driver's personal characteristics.

Method used

A system that acquires user and vehicle characteristics from various services, trains a learning model to predict a driver's tendency using these characteristics, and generates learning data to improve prediction accuracy.

Benefits of technology

Enables prediction of a driver's tendency even without objective data, allowing for personalized insurance policies and improved risk assessment based on individual characteristics.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007583202000001
    Figure 0007583202000001
  • Figure 0007583202000002
    Figure 0007583202000002
  • Figure 0007583202000003
    Figure 0007583202000003
Patent Text Reader

Abstract

To provide a mechanism for predicting a driver of a prescribed tendency based on characteristics of the driver himself / herself.SOLUTION: An information processing device is configured to: acquire user characteristics representing characteristics of a target user from one or more first services; acquire vehicle characteristics representing characteristics relating to a vehicle owned by the target user from a second service different from the one or more first services; and predict whether or not the target user is a driver of a prescribed tendency by using a learning model that has been learned to output a probability that the target user is the driver of the prescribed tendency, with the user tendency and the vehicle characteristics used as inputs.SELECTED DRAWING: Figure 4B
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a driver tendency prediction device, a driver tendency prediction method, a learning device, a learning method, and an information processing program, and in particular to a technique for predicting a driver with a predetermined tendency. [Background technology]

[0002] Drivers of vehicles (including cars and bicycles) are required to drive safely. Drivers who are judged to be able to drive safely are certified as drivers with good driving habits. Furthermore, drivers who are certified as drivers with good driving habits may be eligible for various preferential treatment. For example, if a driver with good driving habits has automobile insurance for the car they own, they will be able to keep the automobile insurance premiums lower than drivers with poor driving habits. Traditionally, whether or not a driver is a good driving habit has been determined based on physical data acquired from the vehicle used and whether or not the driver has had any accidents.

[0003] For example, Patent Document 1 discloses that a driver's driving proficiency is determined by evaluating indices such as the engine revolution rate, the rate of sudden braking, and the rate of sudden steering in the vehicle used by the driver. According to this document, a driver with a high driving proficiency can be determined to be a driver with a good tendency. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2014-65362 A Summary of the Invention [Problem to be solved by the invention]

[0005] As described above, in the past, whether or not a driver has a certain tendency, such as a good tendency, has been determined based on objective data such as physical data obtained by driving and whether or not there has been an accident. However, no mechanism has been proposed so far for predicting whether or not a driver has a certain tendency based on characteristics related to the driver himself, such as the driver's life background, rather than such objective data. If it is possible to predict whether or not a driver has a certain tendency based on characteristics related to the driver himself, it becomes possible to make the prediction even if there is no objective data related to the driver.

[0006] The present invention has been made in consideration of the above-mentioned problems, and has an object to provide a mechanism for predicting drivers with certain tendencies based on characteristics related to the driver himself / herself. [Means for solving the problem]

[0007] In order to solve the above problem, one aspect of a driver tendency prediction device according to the present invention comprises a user characteristic acquisition means for acquiring user characteristics representing characteristics of a target user from one or more first services, a vehicle characteristic acquisition means for acquiring vehicle characteristics representing characteristics related to a vehicle owned by the target user from a second service different from the one or more first services, and a prediction means for predicting whether or not the target user is a driver with a predetermined tendency by using a learning model trained to output the probability that the target user is a driver with a predetermined tendency using the user characteristics and the vehicle characteristics as inputs. In order to solve the above problem, one aspect of a learning device according to the present invention includes a user feature acquisition means for acquiring user features representing the characteristics of a plurality of users from one or more first services, a vehicle feature acquisition means for acquiring vehicle features representing characteristics related to vehicles owned by the plurality of users from a second service different from the one or more first services, a rating acquisition means for acquiring vehicle insurance ratings for the vehicles owned by the plurality of users, a generation means for generating learning data based on the user features, the vehicle features, and the rating, and a learning means for using the learning data to train a learning model for predicting the probability of a driver having a predetermined tendency.

[0008] In order to solve the above problem, one aspect of a driver tendency prediction method according to the present invention includes a user feature acquisition step of acquiring user features representing characteristics of a target user from one or more first services, a vehicle feature acquisition step of acquiring vehicle features representing characteristics related to a vehicle owned by the target user from a second service different from the one or more first services, and a prediction step of predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output the probability that the target user is a driver with a predetermined tendency using the user features and the vehicle features as inputs. In order to solve the above problem, one aspect of a learning method according to the present invention includes a user feature acquisition step of acquiring user features representing characteristics of a plurality of users from one or more first services; a vehicle feature acquisition step of acquiring vehicle features representing characteristics related to vehicles owned by the plurality of users from a second service different from the one or more first services; a rating acquisition step of acquiring vehicle insurance ratings for the vehicles owned by the plurality of users; a generation step of generating learning data based on the user features, the vehicle features, and the rating; and a learning step of using the learning data to train a learning model for predicting the probability of a driver having a predetermined tendency.

[0009] In order to solve the above-mentioned problems, one aspect of the information processing program according to the present invention is an information processing program for causing a computer to execute information processing, the program causing the computer to execute processes including: a user feature acquisition process for acquiring user features representing characteristics of a target user from one or more first services; a vehicle feature acquisition process for acquiring vehicle features representing characteristics related to a vehicle owned by the target user from a second service different from the one or more first services; and a prediction process for predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output the probability that the target user is a driver with a predetermined tendency using the user features and the vehicle features as inputs. Another aspect of the information processing program according to the present invention is an information processing program for causing a computer to execute information processing, the program causing the computer to execute processes including a user feature acquisition process for acquiring user features representing characteristics of a plurality of users from one or more first services, a vehicle feature acquisition process for acquiring vehicle features representing characteristics related to vehicles owned by the plurality of users from a second service different from the one or more first services, a rating acquisition process for acquiring vehicle insurance ratings for the vehicles owned by the plurality of users, a generation process for generating learning data based on the user features, the vehicle features, and the rating, and a learning process for using the learning data to train a learning model for predicting the probability of a driver having a predetermined tendency. Effect of the Invention

[0010] According to the present invention, a mechanism is provided for predicting drivers with certain tendencies based on characteristics related to the driver himself / herself. The above-mentioned objects, aspects, and advantages of the present invention, as well as objects, aspects, and advantages of the present invention not described above, will be understood by those skilled in the art from the following detailed description of the invention by referring to the accompanying drawings and the claims. [Brief description of the drawings]

[0011] [Figure 1] FIG. 1 shows an example of the configuration of an information processing system according to an embodiment. [Diagram 2] FIG. 2 illustrates an example of a functional configuration of an information processing device according to an embodiment. [Figure 3A] FIG. 3A shows a flowchart of the learning procedure of the driver tendency prediction model. [Figure 3B] FIG. 3B shows a conceptual diagram of the driver tendency prediction model during learning. [Figure 4A] FIG. 4A shows a flow chart of the procedure for predicting the drivers of a given trend. [Figure 4B] FIG. 4B shows a conceptual diagram of a driver tendency prediction model when predicting a driver with a predetermined tendency. [Diagram 5] FIG. 5 illustrates an example of the hardware configuration of an information processing device and a user device according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, an embodiment for carrying out the present invention will be described in detail with reference to the attached drawings. Among the components disclosed below, those having the same functions are given the same reference numerals, and their description will be omitted. Note that the embodiment disclosed below is an example of a means for realizing the present invention, and should be appropriately modified or changed depending on the configuration of the device to which the present invention is applied and various conditions, and the present invention is not limited to the following embodiment. In addition, not all of the combinations of features described in the present embodiment are necessarily essential to the solution of the present invention.

[0013] [Information processing system configuration] Fig. 1 shows an example of the configuration of an information processing system according to this embodiment. As an example, as shown in Fig. 1, this information processing system includes an information processing device 10, multiple user devices 11-1 to 11-N (N>1) used by multiple arbitrary users 1 to N, and user device 11-T1 used by user T1. In the following description, unless otherwise specified, user devices 11-1 to 11-N and 11-T1 may be collectively referred to as user device 11. In the following description, the terms user device and user may be used synonymously.

[0014] The user device 11 is, for example, a device such as a smartphone or a tablet, and is configured to be able to communicate with the information processing device 10 via a public network such as LTE (Long Term Evolution) or a wireless communication network such as a wireless LAN (Local Area Network). The user device 11 has a display unit (display surface) such as a liquid crystal display, and each user can perform various operations using a GUI (Graphic User Interface) equipped on the liquid crystal display. The operations include various operations on content such as images displayed on the screen, such as tapping, sliding, and scrolling using a finger or a stylus. Note that user device 11 is not limited to the device shown in Fig. 1, and may be a device such as a desktop PC (Personal Computer) or a notebook PC. In this case, each user may perform operations using input devices such as a mouse and a keyboard. User device 11 may also be provided with a separate display screen.

[0015] A user of the user device 11 can use a plurality of web services (internet-related services) provided from the information processing device 10 or from another device (not shown) via the information processing device 10. These web services can be used by the user, for example, by registering for use with an API (Application Programming Interface) provided by the information processing device 10 or another device.

[0016] In this embodiment, the web service includes a car service. In this embodiment, the car service is a service that provides services related to automobiles. For example, the car service provides the user with information about gas stations and bargain information on car accessories. The car service may also include a service that awards points to users and distributes coupons to users by having the users register information about the automobiles they own. The information about the automobiles may include information such as the automobile manufacturer, model, and replacement frequency.

[0017] Furthermore, in this embodiment, the web service includes an automobile insurance service (vehicle insurance service). The automobile insurance service is a service in which a user who has registered for the service can receive insurance money when an accident or the like occurs with a vehicle owned by the user by paying an amount (insurance premium) according to an assigned grade. The grade is assigned by an insurance company that operates the automobile insurance service, and a high grade is assigned to a user who is determined to be driving safely. Thus, the higher the grade, the higher the discount rate (or the higher the rebate rate), and the lower the amount paid by the user. A predetermined grade is assigned when the service is started to be used, and a high grade can be assigned after a certain period of time has passed by satisfying a certain condition, such as being accident-free for the certain period of time.

[0018] Furthermore, the web services may include online malls and online supermarkets, or services related to communications, finance, real estate, sports, and travel. By using such a web service, the user device 11 can transmit information about the user of the user device 11 to the information processing device 10. For example, the user device 11 can transmit to the information processing device 10 information on characteristics of the user device and the user, such as the IP (Internet Protocol) address of the user device 11, the user's address, and the user's name (first and last name). In addition, the user device 11 performs positioning calculations based on signals received from GPS (Global Positioning System) satellites (not shown), generates the information obtained by the calculations as location information of the user device 11, and transmits it to the information processing device 10.

[0019] In this embodiment, the multiple web services that the user of the user device 11 can use may be directly or indirectly managed and operated by a specific management company and its group companies (affiliated companies of the management company). The multiple web services are configured so that information (e.g., information related to the user, such as user characteristics) can be managed using a user ID that is common among the multiple web services. For example, electronic value such as points that is common among the multiple web services can be granted to the user in response to the user's use of any service among the multiple web services.

[0020] The information processing device 10 acquires various information from the user device 11, learns a learning model based on the information in order to predict a driver with a predetermined tendency, and predicts whether or not a given user is a driver with the predetermined tendency based on the learning model. In this embodiment, the information processing device 10 is configured to learn the learning model using information acquired from the user devices 11-1 to 11-N (i.e., users 1 to N). Furthermore, the information processing device 10 is configured to predict whether or not a user T1 (target user) of the user device 11-T1 is a driver with the predetermined tendency using the learning model. Note that, in this embodiment, the users 1 to N (user devices 11-1 to 11-N) from which information is acquired for learning the learning model (driver tendency prediction model 112 described later) are also referred to as a plurality of sample users.

[0021] [Functional configuration of information processing device 10] 2 shows an example of a functional configuration of the information processing device 10 according to this embodiment. The information processing device 10 can function as a driver tendency prediction device and / or a learning device. The information processing device 10 shown in Fig. 2 includes a user feature acquisition unit 101, a vehicle-related feature acquisition unit 102, a rating data acquisition unit 103, a learning unit 104, a prediction unit 105, an output unit 106, a learning model storage unit 110, and a feature storage unit 120. The learning model storage unit 110 stores a user feature prediction model 111 and a driver tendency prediction model 112. The various learning models will be described later. The feature storage unit 120 stores a user feature 121, a vehicle-related feature 122, and rating data (rating feature) 123. Each of the features will be described later.

[0022] The user feature acquisition unit 101 acquires user features of the user of the user device 11. In this embodiment, the user features are factual features (factual information) about the user device or the user, and are fact-based features (information) that are actually or objectively obtained from the user device or the user. The user feature acquisition unit 101 can acquire the user features directly from the user device 11, for example. In addition, the user feature acquisition unit 101 can acquire the user features as information registered in a predetermined web service by the user of the user device 11. The user features are linked to an identifier of each user (hereinafter, user ID).

[0023] In this embodiment, the user feature acquisition unit 101 acquires, as user features, demographic information registered in one or more web services (one or more first services) and information indicating behavior regarding service usage in the one or more web services (for example, purchase behavior information in past web services). The one or more web services from which user information is acquired may not include car services. The demographic information is information indicating demographic user attributes such as gender, age, residential area, occupation, family structure, etc. The purchase behavior information is, for example, information regarding purchased items (item name, genre, vendor, etc.) and purchase date and time information. In addition, the user feature acquisition unit 101 may acquire, as user features, data indicating preferences registered in a web service by the user at the time of purchase.

[0024] The user feature acquisition unit 101 is also configured to apply the acquired user features to a trained user feature prediction model 111 to acquire user features (estimated user features (attributes)) estimated for the user features. The user feature prediction model 111 is configured to receive user features (i.e., fact features) of an arbitrary user (referred to as a "first user" for the sake of explanation) as input, and to output the probability (correspondence probability) that each of the multiple user features corresponds (matches) the first user. The user feature acquisition unit 101 ultimately determines the estimated user features of the first user from the correspondence probability.

[0025] For example, the user feature acquisition unit 101 inputs data indicating the demographic information of the first user and past purchase history and preferences in web services to the user feature prediction model 111 as the user features of the first user. The user feature prediction model 111 outputs probabilities for each of a plurality of items that the first user is estimated to purchase and a plurality of preferences that the first user may have as the corresponding probability. Then, the user feature acquisition unit 101 can acquire items and preferences having probabilities equal to or greater than a predetermined value as the estimated user features of the first user. Also, the probability itself may be the estimated user feature.

[0026] The estimated user characteristics according to this embodiment may include characteristics related to the user's life stage, such as the probability of working, the probability of having children, etc. Here, the probability of having children may be a probability divided by the child's age, such as the probability of having a child aged 0 to 1, the probability of having a child aged 2 to 6, or the probability of having a child of elementary school age or above. Furthermore, the estimated user characteristics according to this embodiment may include characteristics related to the user's lifestyle, such as the probability of traveling, playing sports, smoking, drinking, subscribing to a newspaper, and purchasing health care products. Here, the probability of traveling may be a probability divided by the purpose or type of travel, such as the probability of traveling alone, traveling domestically, or traveling abroad. Furthermore, the probability of smoking or drinking may be a probability divided by location (e.g., at home or outside (e.g., restaurant)).

[0027] The user feature acquisition unit 101 stores the acquired user features and the estimated user features as user features 121 in the feature storage unit 120. The user features are stored in association with a user ID. The user feature acquisition unit 101 may be configured to store the user features 121 in an external device.

[0028] The car-related feature acquisition unit 102 acquires information registered by the user of the user device 11 in a car service (second service), which is one of the web services, as car-related features (vehicle features). In this embodiment, the car service is a service that provides services related to automobiles, and the user can enjoy the services by registering the car-related features. The car-related features include features such as the manufacturer name, car name, model, contracted automobile insurance company (current and / or past), replacement frequency, and the period since registration with the car service (time information) of the owned automobile. In addition, in this embodiment, the service includes an owned car service (my car service), which is a sub-service included in the car service, and the car-related characteristics may include information on whether or not a user is registered with the owned car service (1 (registered) or 0 (not registered)). In this embodiment, the owned car service is a service that distributes points and coupons free of charge to users who register information about their owned automobiles (e.g., manufacturer name, car name, model). Users can obtain points and coupons free of charge by registering with the owned car service and registering characteristics related to their owned automobiles. Note that users who are registered with the car service do not necessarily have to register with the owned car service.

[0029] The car-related feature is linked to a user ID, and the car-related feature acquisition unit 102 links the acquired car-related feature with the user ID and stores it in the feature storage unit 120 as the car-related feature 122. The car-related feature acquisition unit 102 may be configured to store the car-related feature 122 in an external device.

[0030] The grade data acquisition unit 103 acquires, as grade data, within a legal and appropriate range, a grade assigned by the user of the user device 11 in a car insurance service, which is one of the web services. The grade refers to a classification for determining the amount of insurance premium payment. In one embodiment, the grades are divided into 1st grade to 20th grade in the case of a non-fleet grade that an individual may subscribe to. A high grade is assigned to a user who is determined to be driving safely. Whether or not a user is driving safely is generally determined based on whether or not the user has had an accident with a car owned by the user. A user is assigned a predetermined grade when the user starts using the service, and may be assigned a high grade after the certain period of time has elapsed by satisfying a predetermined condition, such as being accident-free for a certain period of time. The grade data acquisition unit 103 is configured to acquire grade data each time the grade is changed (updated).

[0031] The grade data is linked to a user ID, and the grade data acquiring unit 103 associates the acquired grade data with the user ID and stores it as grade data 123 in the feature storage unit 120. The grade data acquiring unit 103 may be configured to store the grade data 123 in an external device.

[0032] The learning unit 104 generates learning data (teacher data) based on the user features 121, the car-related features 122, and the rating data 123. Furthermore, the learning unit 104 uses the generated learning data to train the driver tendency prediction model 112. The learning procedure of the driver tendency prediction model 112, including the generation of the learning data, will be described later.

[0033] The prediction unit 105 predicts whether or not a given user is a driver with a predetermined tendency, using the driver tendency prediction model 112 learned by the learning unit 104. In this embodiment, the prediction unit 105 predicts whether or not the target user T1 is a driver with a predetermined tendency. A procedure for predicting whether or not a driver has a predetermined tendency will be described later.

[0034] The output unit 106 is configured to output the prediction result by the prediction unit 105. The output may be any output process, and may be output to an external device via a communication I / F (communication I / F 57 in FIG. 5) or may be displayed on a display unit (display unit 56 in FIG. 5). The output unit 106 may generate and output information about the prediction result. For example, when the target user T1 is predicted to be a driver with a predetermined tendency and the target user T1 does not subscribe to an automobile insurance service, the output unit 106 may create and provide an advertisement showing an automobile insurance service to the target user T1.

[0035] [Learning procedure for driver tendency prediction model] The learning procedure of the driver tendency prediction model 112 by the learning unit 104 of this embodiment will be described with reference to Figs. 3A and 3B. Fig. 3A shows a flowchart of the learning procedure of the driver tendency prediction model 112, and Fig. 3B shows a conceptual diagram of the learning of the driver tendency prediction model 112. The process of Fig. 3A can be started after the user feature acquisition unit 101, the car-related feature acquisition unit 102, and the rating data acquisition unit 103 each store the user feature 121, the car-related feature 122, and the rating data 123 in the feature storage unit 120. The user feature 121, the car-related feature 122, and the rating data 123 each include the user feature, the car-related feature, and the rating data of a plurality of sample users (users 1 to N).

[0036] In FIG. 3A, first, the learning unit 104 acquires features of a plurality of sample users (users 1 to N) for learning the driver tendency prediction model 112 (S31 to S33). In S31, the learning unit 104 acquires the user features 121 of a plurality of sample users stored in the feature storage unit 120. In S32, the car-related feature acquisition unit 102 acquires the car-related features 122 of a plurality of sample users stored in the feature storage unit 120. In S33, the grade data acquiring unit 103 acquires the grade data 123 of a plurality of sample users stored in the feature storage unit 120. The processes of S31 to S33 are performed based on the user IDs of the multiple sample users. The processes of S31 to S33 are not limited to the order shown in FIG. 3A, and may be performed in a different order or simultaneously. In the present embodiment, it is assumed that the features acquired in S31 to S33 are stored in the feature storage unit 120 in advance, but this is not limiting. For example, the learning unit 104 may be configured to acquire each of the features directly from at least one of the multiple sample users, or to acquire each of the features from an external device.

[0037] In S34, the learning unit 104 generates (sets) a correct label (correct data) for the driver tendency prediction model 112 from the grade indicated by the grade data acquired in S33. In this embodiment, the predetermined tendency to be predicted is set as a good tendency, and it is defined that a user with a user ID having a grade equal to or higher than the predetermined grade is a driver with a good tendency. In accordance with this, the learning unit 104 generates a correct label (correct data) assuming that the probability that a user with a user ID having a grade equal to or higher than the predetermined grade is a driver with a good tendency is "1" (i.e., the maximum value of the probability). That is, the learning unit 104 generates a correct label of "1" from the grade equal to or higher than the predetermined grade. On the other hand, the learning unit 104 generates a correct label assuming that the probability that a user with a user ID having a grade lower than the predetermined grade is a driver with a good tendency is "0" (i.e., the minimum value of the probability). That is, the learning unit 104 generates a correct label of "0" from the grade lower than the predetermined grade.

[0038] The predetermined grade is set to, for example, 20 grades. Thus, a correct label is generated for a user with grade 20 as a driver with good tendencies (=1), and for a user with grades 1 to 19 as a driver with bad tendencies (=0). The predetermined grade may be input by an operator via an input unit (input unit 55 in FIG. 5) or via a communication unit (communication I / F 57 in FIG. 5). Alternatively, the predetermined grade may be set in advance in the system, or may be set by an arbitrary program stored in a storage unit (ROM 52 or RAM 53 in FIG. 5).

[0039] In the processes up to S34, user features 121, car-related features 122, and correct labels are generated for each of a plurality of sample users (i.e., for each user ID). In the following S35, the learning unit 104 generates data to which the generated correct labels are attached to the user features 121 and the car-related features 122 as learning data. Here, referring to Fig. 3B, the input data 301 includes the user features 121 and the car-related features 122, and the correct labels 302 correspond to the generated correct labels. The learning unit 104 may store the generated learning data in the storage unit (ROM 52 or RAM 53 in Fig. 5).

[0040] In S36, the learning unit 104 uses the learning data generated in S35 to train the driver tendency prediction model 112. Specifically, referring to Fig. 3B, the learning unit 104 inputs input data 301 including user features 121 and vehicle-related features 122 for each of a plurality of sample users to the driver tendency prediction model 112, and trains the driver tendency prediction model 112 so that a correct answer label 302 is output from the driver tendency prediction model 112.

[0041] The driver tendency prediction model 112 is a learning model for machine learning, for example, a learning model based on CatBoost. Alternatively, the driver tendency prediction model 112 may be another boosting-based learning model such as XGBoost or LightGBM. By using a boosting-based learning model, it is also possible to recognize user characteristics attributable to drivers with good driving tendencies.

[0042] The learning unit 104 uses learning data (input data 301 and correct answer label 302) for a plurality of sample users to train the driver tendency prediction model 112. Through training, the driver tendency prediction model 112 is configured to input the user features 121 and car-related features 122 of an arbitrary user and output the probability that the arbitrary user is a driver with good tendencies. When training is completed, in S37, the training unit 104 stores the trained driver tendency prediction model 112 in the training model storage unit 110.

[0043] [Procedure for predicting drivers of a given trend] After the learning unit 104 has learned the driver tendency prediction model 112, the prediction unit 105 uses the learned driver tendency prediction model 112 to predict whether or not the target user T1 (any user) is a driver with a predetermined tendency. In this embodiment, it is predicted whether or not the target user T1 is a driver with a good tendency. The procedure for predicting a driver with a good tendency by the prediction unit 105 in this embodiment will be described with reference to Figs. 4A and 4B. Fig. 4A shows a flowchart of the procedure for predicting a driver with a predetermined tendency, and Fig. 4B shows a conceptual diagram of the driver tendency prediction model 112 when predicting a predetermined driver.

[0044] The process of FIG. 4A is started after the user feature acquisition unit 101 and the car-related feature acquisition unit 102 respectively store the user feature 121 and the car-related feature 122 of the target user T1 in the feature storage unit 120.

[0045] In FIG. 4A, first, the prediction unit 105 acquires input data 401 for the driver tendency prediction model 112 (S41 to S42). In S41, the prediction unit 105 acquires the user features 121 of the target user T1 stored in the feature storage unit 120. In S42, the car-related feature acquisition unit 102 acquires the car-related features 122 of the target user T1 stored in the feature storage unit 120. The processes of S41 to S42 are performed based on the user ID of the target user T1. The processes of S41 to S42 are not limited to the order shown in FIG. 4A, and may be performed in a different order or simultaneously. In the present embodiment, it is assumed that the features acquired in S41 to S42 are stored in the feature storage unit 120 in advance, but this is not limiting. For example, the prediction unit 105 may be configured to acquire the features directly from the target user T1, or to acquire the features from an external device.

[0046] In S43, the prediction unit 105 inputs the user characteristics 121 and the car-related characteristics 122 of the target user T1 acquired in S41 to S42 into the driver tendency prediction model 112, and predicts whether the target user T1 is a driver with a predetermined tendency (in this embodiment, a good tendency). The driver tendency prediction model 112 to which the user characteristics 121 and the car-related characteristics 122 of the target user T1 have been input is configured to output the probability that the target user T1 is a driver with a predetermined tendency (probability 402 of being a driver with a predetermined tendency).

[0047] In S44, the prediction unit 105 transmits to the output unit 106 a prediction result indicating whether the target user T1 is a driver with a predetermined tendency (i.e., a good tendency). For example, the prediction unit 105 may transmit to the output unit 106 the probability that the target user T1 is a driver with a good tendency as a prediction result. Alternatively, a predetermined threshold may be set, and if the probability is equal to or greater than the predetermined threshold, the prediction result indicating that the target user T1 is a driver with a good tendency may be transmitted to the output unit 106, indicating that the target user T1 is a driver with a bad tendency. Note that the predetermined threshold may be input by an operator via an input unit (input unit 55 in FIG. 5) or via a communication unit (communication I / F 57 in FIG. 5). Alternatively, the predetermined threshold may be set in advance in the system, or may be set by an arbitrary program stored in a storage unit (ROM 52 or RAM 53 in FIG. 5).

[0048] The output unit 106 to which the prediction result has been transmitted in S44 can generate information about the prediction result and output it to an external device such as the user device 11-T1. For example, if the prediction result indicates that the target user T1 is a driver with a good tendency and the target user T1 has not subscribed to an automobile insurance service, the output unit 106 may create and provide an advertisement indicating an automobile insurance service to the target user T1. In addition, the output unit 106 can calculate the probability that any multiple users including the target user T1 are a driver with a good tendency from the prediction result for the multiple users, and create a new insurance policy for the automobile insurance service using the probability.

[0049] [Hardware configuration of information processing device 10] FIG. 5 is a block diagram showing an example of a hardware configuration of the information processing device 10 according to the present embodiment. The information processing apparatus 10 according to the present embodiment can be implemented on any single or multiple computers, mobile devices, or any other processing platform. 5, the information processing device 10 is illustrated as being implemented in a single computer, but the information processing device 10 according to the present embodiment may be implemented in a computer system including multiple computers. The multiple computers may be connected to each other via a wired or wireless network so as to be able to communicate with each other.

[0050] 5, the information processing device 10 may include a CPU 51, a ROM 52, a RAM 53, a HDD 54, an input unit 55, a display unit 56, a communication I / F 57, and a system bus 58. The information processing device 10 may also include an external memory. A CPU (Central Processing Unit) 51 generally controls the operations of the information processing device 10, and controls each of the components (52 to 57) via a system bus 58, which is a data transmission path.

[0051] The ROM (Read Only Memory) 52 is a non-volatile memory that stores control programs and the like necessary for the CPU 51 to execute processing. Note that the programs may be stored in a non-volatile memory such as an HDD (Hard Disk Drive) 54 or an SSD (Solid State Drive) or in an external memory such as a removable storage medium (not shown). The RAM (Random Access Memory) 53 is a volatile memory and functions as a main memory, a work area, etc. of the CPU 51. That is, the CPU 51 loads necessary programs, etc. from the ROM 52 to the RAM 53 when executing processing, and realizes various functional operations by executing the programs, etc. The learning model storage unit 110 and the feature storage unit 120 shown in FIG. 2 can be configured by the RAM 53.

[0052] The HDD 54 stores, for example, various data and various information required when the CPU 51 performs processing using a program. The HDD 54 also stores, for example, various data and various information obtained when the CPU 51 performs processing using a program. The input unit 55 is composed of a keyboard and a pointing device such as a mouse. The display unit 56 is configured with a monitor such as a liquid crystal display (LCD), etc. The display unit 56 may be configured in combination with the input unit 55 to function as a GUI (Graphical User Interface).

[0053] The communication I / F 57 is an interface that controls communication between the information processing device 10 and an external device. The communication I / F 57 provides an interface with a network and executes communication with an external device via the network. Various data, various parameters, etc. are transmitted and received between the external device and the communication I / F 57. In this embodiment, the communication I / F 57 may execute communication via a wired LAN (Local Area Network) or a dedicated line conforming to a communication standard such as Ethernet (registered trademark). However, the network available in this embodiment is not limited to this, and may be configured as a wireless network. This wireless network includes wireless PANs (Personal Area Networks) such as Bluetooth (registered trademark), ZigBee (registered trademark), and UWB (Ultra Wide Band). It also includes wireless LANs (Local Area Networks) such as Wi-Fi (Wireless Fidelity) (registered trademark), and wireless MANs (Metropolitan Area Networks) such as WiMAX (registered trademark). It also includes wireless WANs (Wide Area Networks) such as LTE / 3G, 4G, and 5G. Note that the network only needs to connect each device so that they can communicate with each other, and the communication standard, scale, and configuration are not limited to those described above.

[0054] At least some of the functions of each element of the information processing device 10 shown in Fig. 2 can be realized by the CPU 51 executing a program. However, at least some of the functions of each element of the information processing device 10 shown in Fig. 2 may be operated as dedicated hardware. In this case, the dedicated hardware operates under the control of the CPU 51.

[0055] [Hardware configuration of user device 11] The hardware configuration of the user device 11 shown in Fig. 1 may be the same as that of Fig. 5. That is, the user device 11 may include a CPU 51, a ROM 52, a RAM 53, a HDD 54, an input unit 55, a display unit 56, a communication I / F 57, and a system bus 58. The user device 11 can display various information provided by the information processing device 10 on the display unit 56, and perform processing corresponding to an input operation received from a user via a GUI (configured by the input unit 55 and the display unit 56).

[0056] In this way, the information processing device 10 can predict whether the target user is a driver with a predetermined tendency based on the user characteristics of the target user obtained through the web service and information related to the car owned by the target user (car-related characteristics). This makes it possible to predict whether the target user is a driver with a predetermined tendency even if there is no objective data on the target user, such as physical data obtained by driving or whether there have been any accidents.

[0057] Furthermore, the information processing device 10 can provide various new services to users based on the prediction results of drivers with a certain tendency. For example, the information processing device 10 can predict that a target user is a driver with a good tendency, and when the target user does not subscribe to a specific automobile insurance service, can recommend the specific automobile insurance service to the target user. Furthermore, the information processing device 10 can calculate the probability that any multiple users including the target user are a driver with a good tendency from the prediction results for the multiple users, and create a new insurance policy for the automobile insurance service using the probability. This can improve the profitability of the automobile insurance service.

[0058] In addition, in this embodiment, for the prediction, a driver tendency prediction model 112, which is a learning model for machine learning, is constructed. Here, the driver tendency prediction model 112 is trained using training data generated from user features, car-related features, and grades in the automobile insurance service obtained from multiple users. Since the training data includes correct answer labels created based on grades actually assigned in the automobile insurance service, the driver tendency prediction model 112 is trained to output more accurate prediction results by taking into account the background of the grade assignment.

[0059] In the above embodiment, an embodiment related to an automobile has been described, but the above embodiment can be applied to any vehicle. For example, the above embodiment can be applied to other vehicles such as bicycles and motorcycles.

[0060] In the above embodiment, the driver tendency prediction model 112 is configured to receive the user characteristics 121 and the vehicle-related characteristics 122 of the target user as input and output the probability that the target user is a driver with a predetermined tendency, as shown in Fig. 3B and Fig. 4B, but the input data is not limited to this. For example, the driver tendency prediction model 112 may be configured and trained to receive only the user characteristics 121 of the target user as input and output the probability that the target user is a driver with a predetermined tendency.

[0061] Although specific embodiments have been described above, the embodiments are merely illustrative and are not intended to limit the scope of the present invention. The apparatus and method described in this specification can be embodied in forms other than those described above. Furthermore, the above-described embodiments can be appropriately omitted, substituted, and modified without departing from the scope of the present invention. Such omitted, substituted, and modified forms are included in the scope of the claims and their equivalents, and belong to the technical scope of the present invention.

[0062] The disclosure of this embodiment includes the following configuration. [1] A driver tendency prediction device comprising: a user feature acquisition means for acquiring user features representing characteristics of a target user from one or more first services; a vehicle feature acquisition means for acquiring vehicle features representing characteristics related to a vehicle owned by the target user from a second service different from the one or more first services; and a prediction means for predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output the probability that the target user is a driver with a predetermined tendency using the user features and the vehicle features as inputs.

[0063] [2] The driver tendency prediction device described in [1], wherein the user characteristics include factual characteristics about the target user and estimated characteristics estimated from the factual characteristics.

[0064] [3] The driver tendency prediction device described in [2], wherein the fact characteristics include demographic information of the target user registered in the one or more first services and information indicating behavior regarding service usage in the first service.

[0065] [4] The driver tendency prediction device according to [2] or [3], wherein the estimated characteristics include at least one of characteristics related to the target user's life stage and characteristics related to the target user's lifestyle.

[0066] [5] A driver tendency prediction device according to any one of [1] to [4], wherein a plurality of services including the one or more first services and the second service are configured so that information can be managed among the plurality of services using a common user identifier.

[0067] [6] A user feature acquisition means for acquiring user features representing characteristics of a plurality of users from one or more first services, and a vehicle feature acquisition means for acquiring vehicle features representing characteristics related to vehicles owned by the plurality of users from a second service different from the one or more first services; A learning device comprising: a rating acquisition means for acquiring vehicle insurance ratings for vehicles owned by the multiple users; a generation means for generating learning data based on the user characteristics, the vehicle characteristics, and the rating; and a learning means for using the learning data to train a learning model for predicting the probability of a driver having a predetermined tendency.

[0068] [7] The learning device described in [6], wherein the user features include factual features for the multiple users and estimated features estimated from the factual features.

[0069] [8] The learning device described in [7], wherein the fact characteristics include demographic information of the target user registered in the one or more first services and information indicating behavior regarding service usage in the first service.

[0070] [9] The learning device described in [7] or [8], wherein the estimated features include at least one of features related to the life stages of the multiple users and features related to the lifestyles of the multiple users.

[0071]

[10] A learning device described in any of [6] to [9], wherein a plurality of services including one or more first services and the second service are configured so that information can be managed between the plurality of services using a common user identifier. [Explanation of symbols]

[0072] 1 to N: users, T1: target user, 10: information processing device, 11-1 to 11-N, 11-T1: user device, 101: user feature acquisition unit, 102: vehicle-related feature acquisition unit, 103: rating data acquisition unit, 104: learning unit, 105: prediction unit, 106: output unit, 110: learning model storage unit, 111: user feature prediction model, 112: driver tendency prediction model, 120: feature storage unit, 121: user feature, 122: vehicle-related feature, 123: rating data

Claims

1. A fact feature acquisition means for acquiring fact features about a target user from one or more services; an estimated feature acquisition means for acquiring estimated features of the target user by applying the fact features to a trained user feature prediction model; A prediction means for predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output a probability that the target user is a driver with a predetermined tendency, using the fact feature and the estimated feature as input; A driver tendency prediction device comprising:

2. The fact characteristics include demographic information of the target user registered in the one or more first services and information indicating behavior regarding service usage in the first service; 2. The driver tendency prediction device according to claim 1.

3. The estimated characteristics include at least one of characteristics related to a life stage of the target user and characteristics related to a lifestyle of the target user.

3. The driver tendency prediction device according to claim 1 or 2.

4. The plurality of services including the one or more services are configured so that information can be managed by a common user identifier among the plurality of services.

2. The driver tendency prediction device according to claim 1.

5. The user feature prediction model is configured to receive the fact feature as an input and output a probability that each of a plurality of user features corresponds to the target user; The estimated feature acquisition means acquires the estimated feature based on the probability of occurrence.

2. The driver tendency prediction device according to claim 1.

6. A fact feature acquisition means for acquiring fact features about a plurality of users from one or more services; an estimated feature acquisition means for acquiring estimated features estimated for the plurality of users by applying the actual features to a trained user feature prediction model; A rating acquisition means for acquiring vehicle insurance ratings for vehicles owned by the plurality of users; A generating means for generating learning data based on the factual features, the estimated features, and the grades; A learning means for learning a learning model for predicting the probability of a driver having a predetermined tendency, using the learning data; A learning device comprising:

7. The fact characteristics include demographic information of the plurality of users registered in the one or more first services and information indicating behavior regarding service usage in the first service; 7. The learning device according to claim 6.

8. The estimated characteristics include at least one of characteristics related to life stages of the plurality of users and characteristics related to lifestyles of the plurality of users.

8. The learning device according to claim 6 or 7.

9. The plurality of services including the one or more services are configured so that information can be managed by a common user identifier among the plurality of services.

7. The learning device according to claim 6.

10. The user feature prediction model is configured to receive the fact feature as an input and output a probability that each of a plurality of user features corresponds to the plurality of users; The estimated feature acquisition means acquires the estimated feature based on the probability of occurrence.

7. The learning device according to claim 6.

11. A driver tendency prediction method executed by an information processing device, comprising: A fact feature acquisition step of acquiring fact features about a target user from one or more services; an estimated feature acquisition step of acquiring estimated features of the target user by applying the fact features to a trained user feature prediction model; a prediction step of predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output a probability that the target user is a driver with a predetermined tendency using the factual features and the estimated features as inputs; A driver tendency prediction method comprising the steps of:

12. A learning method executed by an information processing device, comprising: A fact feature acquisition step of acquiring fact features for a plurality of users from one or more services; an estimated feature acquisition step of acquiring estimated features estimated for the plurality of users by applying the actual features to a trained user feature prediction model; A grade acquisition step of acquiring vehicle insurance grades for vehicles owned by the plurality of users; A generation step of generating learning data based on the factual features, the estimated features, and the grades; A learning process of learning a learning model for predicting the probability of a driver having a predetermined tendency, using the learning data; A learning method comprising:

13. An information processing program for causing a computer to execute information processing, the program comprising: A fact feature acquisition process for acquiring fact features about a target user from one or more services; an estimated feature acquisition process for acquiring estimated features of the target user by applying the fact features to a trained user feature prediction model; a prediction process for predicting whether the target user is a driver with a predetermined tendency by using a learning model trained to output a probability that the target user is a driver with a predetermined tendency, using the fact feature and the estimated feature as inputs; Information processing program.

14. An information processing program for causing a computer to execute information processing, the program comprising: a fact feature acquisition process for acquiring fact features for a plurality of users from one or more services; an estimated feature acquisition process for acquiring estimated features of the plurality of users by applying the actual features to a trained user feature prediction model; A grade acquisition process for acquiring vehicle insurance grades for vehicles owned by the plurality of users; A generation process for generating learning data based on the fact features, the estimated features, and the grades; A learning process for learning a learning model for predicting the probability of a driver having a predetermined tendency, using the learning data. Information processing program.

Citation Information

Patent Citations

  • Online system and method for insurance underwriting

    JP2013502651A

  • Service provision system, information provision device, service provision method, and program

    JP2014065362A

  • Individualized adaptation of driver action prediction models

    JP2018027776A

  • Information processing device, classification method and computer program

    JP2019093896A

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

    JP2020190778A