JP7576058B2 - Information processing system, method and program - Google Patents

Description

This disclosure relates to technology for supporting evaluations such as calculating scores for users.

Conventionally, a judgment device has been proposed that includes a user information acquisition unit that acquires behavioral information indicating a user's behavior, and a credit judgment unit that judges the user's creditworthiness regarding his or her future ability to repay loans based on the behavioral information (see Patent Document 1). Also proposed is a credit score management server that includes an input information acquisition unit that acquires parameter information entered into a format displayed on a user terminal, a behavior history acquisition unit that acquires the user's behavior history continuously transmitted from the user terminal, and a credit score calculation unit that calculates a credit score based on the parameters and behavior history (see, for example, Patent Document 2).

JP 2021-174039 A JP 2020-009226 A

Conventionally, technologies have been proposed that calculate a user score that represents a user's credibility, etc., based on the user's behavioral history. In addition, various technologies have been proposed that improve the convenience of services provided to users by utilizing geographic locations related to the user, such as the user's place of residence. However, when calculating the user score, information directly related to the target user is mainly treated as the attributes of that user, and there is room for improvement in improving the accuracy of the user score.

In view of the above problems, the present disclosure aims to realize information processing for reflecting the context inferred from the geographic location of a user in the user's evaluation.

An example of the present disclosure is an information processing system including an attribute determination means for determining a geographic fact attribute related to a geographic location based on a geographic location related to a user and predetermined map data, an attribute estimation means for estimating at least a part of a group of attribute data related to the user based on the determined geographic fact attribute, and a user score estimation means for estimating a user score to be set for the user based on the estimated group of attribute data.

The present disclosure can be understood as an information processing device, a system, a method executed by a computer, or a program executed by a computer. The present disclosure can also be understood as such a program recorded on a recording medium readable by a computer or other device, machine, etc. Here, a recording medium readable by a computer, etc. refers to a recording medium that stores information such as data and programs through electrical, magnetic, optical, mechanical, or chemical action and can be read by a computer, etc.

According to the present disclosure, it is possible to realize information processing for reflecting the context inferred from the geographical location of a user in the user's evaluation.

1 is a schematic diagram showing a configuration of an information processing system according to an embodiment. 1 is a diagram illustrating an outline of a functional configuration of an information processing device according to an embodiment. FIG. 2 is a simplified diagram of a geographical attribute estimation process according to an embodiment. 1 is a simplified diagram of a user score estimation process according to an embodiment. FIG. 1 is a simplified diagram illustrating the concept of a decision tree of a machine learning model employed in an embodiment. 1 is a flowchart illustrating a flow of a machine learning process according to an embodiment. 11 is a flowchart showing a flow of a user score estimation process according to the embodiment. FIG. 13 is a diagram illustrating an outline of a functional configuration of an information processing device according to a first variation. 13 is a simplified diagram of a user score estimation process according to a second variation. 13 is a simplified diagram of a user score estimation process according to a third variation. 13 is a simplified diagram of a user score estimation process according to a fourth variation. 13 is a simplified diagram of a user score estimation process according to a fifth variation.

Below, an embodiment of an information processing device, method, and program according to the present disclosure will be described with reference to the drawings. However, the embodiment described below is merely an example of an embodiment, and the information processing device, method, and program according to the present disclosure are not limited to the specific configurations described below. In implementing the present invention, a specific configuration according to the mode of implementation may be appropriately adopted, and various improvements and modifications may be made. The present invention allows at least a portion of the configurations in the embodiments and variations described below to be adopted in each other as appropriate.

In this embodiment, an embodiment will be described in which the technology according to the present disclosure is implemented for a user score management system that manages a user score that indicates some measure related to a user (e.g., credit, etc.). However, the technology according to the present disclosure can be widely used for technology to support user evaluation, such as estimating user scores, and the application of the present disclosure is not limited to the examples shown in the embodiment.

<System Configuration>
1 is a schematic diagram showing the configuration of an information processing system according to this embodiment. In the system according to this embodiment, an information processing device 1 and one or more service providing systems 5 are connected to each other so as to be able to communicate with each other. A user is a user of a service provided by the service providing system 5, and receives the service by accessing the service providing system 5 from a user terminal.

The information processing device 1 is a computer including a CPU (Central Processing Unit) 11, a ROM (Read Only Memory) 12, a RAM (Random Access Memory) 13, a storage device 14 such as an EEPROM (Electrically Erasable and Programmable Read Only Memory) or a HDD (Hard Disk Drive), a communication unit 15 such as a NIC (Network Interface Card), etc. However, the specific hardware configuration of the information processing device 1 can be omitted, replaced, or added as appropriate depending on the embodiment. Furthermore, the information processing device 1 is not limited to a device consisting of a single housing. The information processing device 1 may be realized by multiple devices using so-called cloud or distributed computing technology, etc.

The information processing device 1 manages a user score for each user and provides the user score to the service providing system 5. The service providing system 5 can customize services for the target user according to the user score provided by the information processing device 1.

The service providing system 5 is a computer equipped with a CPU, ROM, RAM, a storage device, a communication unit, an input device, an output device, etc. (not shown). Furthermore, these systems and terminals are not limited to devices consisting of a single housing. These systems and terminals may be realized by multiple devices using so-called cloud or distributed computing technology, etc.

The services provided by the service providing system 5 are, for example, map information services, credit card/postpaid payment services, electronic money payment services, online shopping services, online reservation services, operation center services, etc. Note that "postpaid payment" is not limited to services such as Buy Now Pay Later (BNPL), but includes any purchase of products/services with postpaid payment. The services provided by the service providing system 5 are not limited to the examples in this embodiment. The service providing system 5 notifies the information processing device 1 of the attribute data group 221 of the user acquired from the user when the service is provided. Here, the attribute data group 221 of the user includes service usage history data by the user. The content of the service usage history data varies depending on the content of the service, and may include, for example, history data of the user's location information, payment history data of credit card usage amount/postpaid payment usage amount, electronic money usage history data, transaction history data, reservation history data, operation history data for the user from the operation center, etc.

FIG. 2 is a diagram showing an outline of the functional configuration of the information processing device 1 according to this embodiment. The information processing device 1 functions as an information processing device including an attribute determination unit 21, an attribute estimation unit 22, a user score estimation unit 23, and a machine learning unit 24, by a program recorded in the storage device 14 being read into the RAM 13 and executed by the CPU 11, which controls each piece of hardware included in the information processing device 1. Note that in this embodiment and other embodiments described below, each function included in the information processing device 1 is executed by the CPU 11, which is a general-purpose processor, but some or all of these functions may be executed by one or more dedicated processors.

The attribute determination unit 21 determines the geographic fact attribute 211 related to the geographic location of the user based on the geographic location of the user and the predetermined map data. Examples of the geographic location of the user include the user's place of residence, place of work, place of study, etc. Other geographic locations of the user may include, for example, commercial areas, recreational areas, tourist spots, etc. that the user frequently visits, and the types of geographic locations of the user are not limited to the examples in this embodiment. In this embodiment, the attribute determination unit 21 estimates the geographic location of the user based on the user's behavioral information (behavioral history), and determines the geographic fact attribute 211 related to the geographic location based on the estimated geographic location of the user and the predetermined map data. Here, the information used as the user's behavioral information (behavioral history) is, for example, history data of the user's location information that can be obtained from the service provision system 5, etc. Here, the technology for acquiring user location information by the service providing system 5 and the like and the type of location information are not limited, and it is possible to use location information of the user terminal acquired by GNSS such as GPS, transmission and reception of signals between the user terminal and base stations such as mobile phone base stations and satellite base stations, location information of the user terminal estimated based on beacons from beacon sources such as Wi-Fi access points, etc. Here, the beacon may be, for example, a physical beacon, a virtual beacon, a beacon that can be received from an indoor beacon source, a beacon that can be received from an outdoor beacon source, or a combination thereof.

In this embodiment, the attribute determination unit 21 estimates a geographical location related to the user based on the user's behavioral information (behavioral history). For example, the attribute determination unit 21 can aggregate historical data of location information accumulated for the target user and estimate the geographical location related to the user based on the length of time the user stayed and the time period in which the user stayed. More specifically, when it is desired to estimate the user's residence as the geographical location related to the user, the attribute determination unit 21 can estimate, as the user's residence, a location where the user's location information is concentrated during a predetermined time period (for example, the nighttime period from 18:00 to 8:00) in which the user is generally likely to be at home. In addition, when it is desired to estimate the user's place of work or place of school, etc. as the geographical location related to the user, the attribute determination unit 21 can estimate, as the user's place of work, etc., a location where the user's location information is concentrated during a predetermined time period (for example, the daytime period from 8:00 to 18:00) in which the user is generally likely to be working or studying. Note that, here, an example is described in which the location where the user is staying during the nighttime is estimated as the residence, and the location where the user is staying during the daytime is estimated as the workplace, etc., but the time period referenced to estimate the user's residence and/or workplace, etc. is not limited to a specific time period. The time period referenced to estimate the user's residence and/or workplace, etc. may be changed depending on the user's occupation and working hours obtained by referring to other attribute data.

The attribute determination unit 21 may also estimate the geographical location and the type of the geographical location related to the user according to the geographical fact attribute 211 of the place where the user stays. That is, if the location where the target user stays for a predetermined time or more has the geographical fact attribute 211 of a residential area, the location may be estimated as the user's residence, and if the location where the target user stays for a predetermined time or more has the geographical fact attribute 211 of an office district, the location may be estimated as the user's workplace, etc. In this case, the attribute determination unit 21 can estimate the type of geographical location related to the user by referring to map data that can determine the type of geographical fact attribute 211, such as whether the specified location is a residential area or an office district.

In this embodiment, a method is adopted in which the geographical location of the user is estimated based on the user's behavioral information (behavioral history), but the geographical location of the user may be obtained by other methods. For example, address information, workplace information, and school information registered by the user themselves may be used as the geographical location of the user. In addition, the user's behavioral history may be a user's behavioral history obtained using a login terminal or a camera installed at a predetermined location, a user's behavioral history updated in conjunction with a payment process performed by the user with the presentation of a point card associated with a user identifier at a predetermined location, a user's behavioral history updated in conjunction with a payment process performed by a user terminal carried by the user at a predetermined location, a user's behavioral history updated in conjunction with a code image reading process installed at a predetermined location by a user terminal carried by the user when going out, or a user's behavioral history updated in conjunction with near-field wireless communication (NFC) performed between a contactless reader installed at a predetermined location and a user terminal carried by the user when going out.

When the geographical position of the user is estimated, the attribute determination unit 21 determines the geographical fact attribute 211 related to the geographical position based on the predetermined map data. Here, the geographical fact attribute 211 is information indicating whether or not a certain geographical position has some attribute (presence or absence of an attribute, attribute value). The geographical fact attribute 211 is appropriately used for estimating a geographical score indicating the probability that a predetermined attribute (geographical estimated attribute) is applied to the geographical position. For this reason, the map data is not limited to general map data as long as it is data that can acquire the geographical fact attribute 211 corresponding to the position by specifying the geographical position. In addition, the method of specifying the geographical position is generally latitude and longitude, but an address or postal code indicating a city, ward, town, or village may also be used, and the specific method of specification and format of specification are not limited. As the map data, data such as the following examples can be used.

(1) Map data capable of identifying a past state The map data may be data capable of identifying a past state of a geographical location as a geographical fact attribute 211 corresponding to the geographical location. In this case, the attribute determination unit 21 determines the geographical fact attribute 211 related to the geographical location where the target user may be located based on the past state corresponding to the geographical location related to the target user. For example, examples of past states corresponding to a geographical location include an old river and an old three-industry area, and these past states may be related to future land prices, etc., and therefore may be treated as geographical fact attributes 211 for some kind of estimation processing, etc. In addition, for example, examples of past states corresponding to a geographical location include the sea before reclamation, and these past states may be related to future land prices, etc., and therefore may be treated as geographical fact attributes 211.

(2) Map data capable of identifying a relationship between a geographical location and other geographical locations existing in the vicinity The map data may be data capable of identifying a relationship between the location and other geographical locations existing in the vicinity of the location as a geographical fact attribute 211 corresponding to the geographical location. In this case, the attribute determination unit 21 determines the geographical fact attribute 211 related to the geographical location where the target user may be located based on the relationship between the geographical location related to the target user and other geographical locations existing in the vicinity of the geographical location. For example, the relationship between a geographical location and other geographical locations existing in the vicinity of the location may include an elevation difference (high/low) compared to the surroundings, a high or low surface ground amplification rate/a high or low surface ground amplification rate, a distance from a first-class river, a distance from a shrine or temple established more than 400 years ago, and the presence or absence of a shrine or temple established more than 400 years ago within 1 km. These relationships may be related to future land prices, disaster risks, etc., and therefore may be treated as the geographical fact attribute 211. Furthermore, for example, the relationship between a geographical location and other geographical locations existing around the location includes the presence or absence of a station, hospital, park, government office, library, etc., in the vicinity of the geographical location, and these relationships may be related to future land prices, real estate prices, etc., and therefore may be treated as the geographical fact attribute 211. Furthermore, for example, the relationship between a geographical location and other geographical locations existing around the location includes the presence or absence of a corporate base such as a company building, factory, manufacturing base, etc., in the vicinity of the geographical location, and these relationships may be related to future land prices, tax revenues, etc., and therefore may be treated as the geographical fact attribute 211. Furthermore, for example, the relationship between a geographical location and other geographical locations existing around the location includes the presence or absence of a pachinko parlor, a horse racing track, a boat racing track, a bicycle racing track, a public racing track such as an auto race track, etc., in the vicinity of the geographical location, and these relationships may be related to future land prices, etc., and therefore may be treated as the geographical fact attribute 211.

(3) Map data capable of identifying information related to a region The map data may be data capable of identifying information related to a region including a geographical position as a geographical fact attribute 211 corresponding to the geographical position. In this case, the attribute determination unit 21 determines the geographical fact attribute 211 related to the geographical position where the target user may be located based on information related to the region including the geographical position related to the target user. For example, information related to the region including the geographical position includes statistical information on public safety and economic indicators, and such regional information may be related to future land prices, etc., and therefore may be treated as the geographical fact attribute 211 for some kind of estimation processing, etc. In addition, for example, information related to the region including the geographical position includes the total population, population increase/decrease, future population estimate, population by age, population by marital status, population by Japanese/foreigner, household situation (proportion by family type, trend in number of general households by family type), housing situation (trend in number of general households living in houses by construction type), basic population tabulation results such as foreigner population by nationality, etc., which are revealed by the census, and such regional information may be related to future land prices, etc., and therefore may be treated as the geographical fact attribute 211. Furthermore, for example, information relating to an area including a geographical position includes basic aggregate results of employment status and the like revealed by a national census, such as the labor force status of the population, and the composition of industrial and occupational major classifications such as married couples and households with children, and since this regional information may be related to future land prices, etc., it can be treated as the geographical fact attribute 211. Furthermore, for example, information relating to an area including a geographical position includes national land information such as people flow data, and since this regional information may be related to future land prices, it can be treated as the geographical fact attribute 211.

3 is a simplified diagram of the estimation process of the geographical estimated attribute according to the embodiment. The attribute estimation unit 22 estimates the attribute data group 221 related to the user. The attribute data group 221 related to the user includes attribute data including the usage history data acquired from the service providing system 5. At this time, the attribute estimation unit 22 estimates the geographical estimated attribute 222 as at least a part of the attribute data group 221 related to the user based on the geographical fact attribute 211 determined by the attribute determination unit 21. The attribute estimation unit 22 may estimate the geographical estimated attribute 222 based on one or more determined geographical fact attributes 211 and determine the geographical estimated attribute 222 as at least a part of the attribute data group 221 related to the user. In this embodiment, the attribute estimation unit 22 determines the geographical estimated attribute 222 as a part of the attribute data group 221 related to the user as it is, and uses it as the attribute data group 221. The attribute estimation unit 22 may perform some processing (normalization, ranking, labeling, etc.) on the determined geographic fact attribute 211 to make it a part of the attribute data group 221, or may use the determined geographic fact attribute 211 to calculate other types of scores or labels as attribute data. Here, other machine learning models may be involved in the calculation of other types of scores or labels. For example, the attribute estimation unit 22 may determine at least a part of the attribute data group 221 related to the user using an attribute estimation model 220 generated using teacher data in which the determined geographic fact attribute 211 is an input value and the geographic score 222A or the geographic label 222B is an output value. Here, the geographic score 222A is a score indicating the degree to which a predetermined attribute and attribute value (geographic estimation attribute) can be applied, and the geographic label 222B is information indicating the type of the predetermined attribute and attribute value.

The geographic score 222A included in the geographic estimation attribute 222 is, for example, a score indicating the appropriateness (probability) of applying the geographic label 222B, which indicates that the area is an area where other users in a predetermined situation are located for more than a predetermined time, to the area indicated by the geographic location of the user. The geographic score 222A is, for example, a score indicating the appropriateness of applying the geographic label 222B, which indicates that the area is an area where other users in a predetermined situation are located more than a predetermined percentage, to the area indicated by the geographic location of the user. The predetermined situation here may be, for example, a living situation based on a life stage, lifestyle, etc., or an economic situation based on income, assets, etc. The geographic score 222A included in the geographic estimation attribute 222 is, for example, a score indicating the appropriateness (probability) of applying the geographic label 222B, which indicates that the area is an area where other users may be in a predetermined situation, to the area indicated by the geographic location of the user. The predetermined situation here may be a situation in which assets are damaged by a disaster, etc. The geographic score 222A included in the geographic estimation attribute 222 is, for example, a score indicating the appropriateness (likelihood) of applying the geographic label 222B corresponding to the geographic trend to the area indicated by the geographic location of the user. The geographic trend here may be a situation in which asset values increase due to redevelopment, etc., a situation in which the standard of living increases due to rising prices, etc., or a situation in which consumption of goods and services becomes active due to the concentration of people, etc.

Here, the attribute data group 221 estimated by the attribute estimation unit 22 may include demographic attributes 223, behavioral attributes 224, or psychographic attributes 225. The demographic attributes 223 may be, for example, the user's gender, family structure, age, etc., and the behavioral attributes 224 may be based on service usage history data, such as whether or not a cash advance has been used, whether or not a revolving payment has been used, a deposit and withdrawal history for a specific account, a commercial transaction history for some product including gambling or lottery (which may include an online transaction history in an online marketplace, etc.), etc., and the psychographic attributes 225 may be, for example, a preference for gambling or lottery. However, the available user attributes are not limited to the examples in this embodiment. For example, "time required for operation (calls, etc.)" and "credit card usage amount/deferred payment usage amount" from an operation center service, etc. may also be used as attributes.

The attribute data group 221 further includes geographical estimated attributes 222 estimated based on the geographical fact attributes 211 determined by the attribute determination unit 21 in addition to the attributes described above. The geographical estimated attributes 222 determined based on the geographical fact attributes 211 may include, for example, residence attribute data, workplace attribute data, and/or school attribute data expressed in a data format such as a geographical score 222A (e.g., a continuous value between 0 and 1) or a geographical label 222B (e.g., a binary value corresponding to presence/absence or yes/no).

Figure 4 is a simplified diagram of a user score estimation process according to an embodiment. The user score estimation unit 23 estimates a user score 231 to be set for a user based on the attribute data group 221. In this embodiment, the user score estimation unit 23 estimates (calculates) the user score 231 to be set for the user by inputting the attribute data group 221 of the user to a user score estimation model 230. Here, the output value of the user score estimation model 230 is, for example, a user score normalized/standardized with 0 as the minimum value and 1 as the maximum value.

The machine learning unit 24 generates and/or updates a user score estimation model 230 used for user score estimation by the user score estimation unit 23. The machine learning model for user score estimation is a machine learning model that outputs a user score 231 indicating some measure related to the user (e.g., trust, etc.) when one or more attribute data (attribute data group 221) related to a target user is input.

When generating and/or updating the user score estimation model 230, the machine learning unit 24 creates teacher data for each user based on the data acquired from the service providing system 5, in which the attribute data group 221 including the demographic attributes of the user is defined as an input value and the user score 231 related to the user is defined as an output value. Then, the machine learning unit 24 creates the user score estimation model 230 based on the teacher data. As described above, the attribute data group 221 input to the user score estimation model 230 includes attribute data estimated based on the geographic fact attribute 211 determined by the attribute determination unit 21, and is combined with the user score 231 of the corresponding user and input to the machine learning unit 24 as teacher data. The user score 231 set in the teacher data may be a user score 231 determined based on a rule base, or may be a user score 231 set manually (annotated). It may also be a user score 231 corrected by an administrator or the like after being output in the past by the user score estimation model 230.

A machine learning model generation framework that can be adopted as the user score estimation model 230 or the like when implementing the technology according to the present disclosure is based on an ensemble learning algorithm, for example. For example, a machine learning framework (for example, LightGBM) based on a gradient boosting decision tree (GBDT) may be adopted as the framework. In other words, the framework may be a machine learning framework based on a decision tree model that transfers the error between the correct answer and the predicted value between the previous and next weak learners (weak classifiers). The predicted value here refers to the predicted value of the user score 231, for example. In addition to LightGBM, the framework may adopt boosting methods such as XGBoost and CatBoost. According to a framework using a decision tree, a machine learning model with relatively high performance can be generated with less effort in parameter adjustment compared to a framework using a neural network. However, the framework for generating a machine learning model that can be used to implement the technology disclosed herein is not limited to the example in this embodiment. For example, instead of a gradient boosting decision tree, another learning device such as a random forest may be used as a learning device, or a learning device that is not called a weak learning device such as a neural network may be used. In particular, when a learning device that is not called a weak learning device such as a neural network is used, ensemble learning does not need to be used.

5 is a simplified diagram showing the concept of a decision tree of a machine learning model employed as the user score estimation model 230 in this embodiment. When a gradient boosting machine learning framework based on a decision tree algorithm is employed, the branching conditions of each node of the decision tree are optimized. Specifically, in the gradient boosting machine learning framework based on a decision tree algorithm, the user scores 231 are calculated for each user group having attributes indicated by each of the two child nodes branched from one parent node, and the branching conditions of the parent node are optimized so that the difference between the user scores 231 is large (for example, so that the difference is maximized or is equal to or greater than a predetermined threshold), that is, so that the two child nodes branch neatly. For example, if the attribute indicated as the branching condition of the node is a geographic score, the predetermined score set as the branching threshold may be changed, or the branching condition may be changed to something other than the geographic score. Also, for example, if the attribute indicated as the branching condition of the node is age, the age set as the branching threshold may be changed, or the branching condition may be changed to an attribute other than age. In this way, by recursively optimizing the branching conditions of all nodes in the decision tree, the estimation accuracy of the user score 231 based on the attribute data group 221 can be improved.

When the attribute estimation unit 22 estimates at least a part of the attribute data group 221 using the attribute estimation model 220, the machine learning unit 24 further generates and/or updates the attribute estimation model 220 used by the attribute estimation unit 22 to determine the geographical estimated attribute 222 based on the geographical fact attribute 211. The attribute estimation model 220 is a machine learning model that outputs a geographical estimated attribute 222 related to the geographical location where the user is located when one or more geographical fact attributes 211 related to the target user are input.

In generating and/or updating the attribute estimation model 220, the machine learning unit 24 creates teacher data in which one or more geographic fact attributes 211 are defined as input values and one geographic estimation attribute 222 is defined as an output value based on the data acquired from the service providing system 5. Then, the machine learning unit 24 generates or updates the attribute estimation model 220 based on the teacher data. One or more geographic estimation attributes 222 are combined with corresponding attribute data (geographic fact attributes 211) and input to the machine learning unit 24 as teacher data. In addition, in generating or updating the attribute estimation model 220, a machine learning model generation framework that can be adopted is not limited, but a machine learning framework of gradient boosting based on a decision tree algorithm may be adopted, as in the user score estimation model 230 described above. In this way, by obtaining an output value (geographic estimation attribute 222) based on one or more geographic fact attributes 211 using the attribute estimation model 220, the output value implicitly reflects the context related to the user's place of residence, place of work, place of study, etc. Here, the attribute estimation model 220 may be trained with data labeled with the amount of geographical score 222A, or may be trained separately with data labeled with the presence or absence of a specific geographical label 222B.

<Processing flow>
Next, a process flow executed by the information processing system according to the present embodiment will be described. Note that the specific contents and processing order of the processes described below are an example for implementing the present disclosure. The specific contents and processing order may be appropriately selected according to the embodiment of the present disclosure.

Figure 6 is a flowchart showing the flow of the machine learning process according to this embodiment. The process shown in this flowchart is executed at a timing specified by the administrator.

In this embodiment, in the machine learning process, the user score estimation model 230 is generated and/or updated. The machine learning unit 24 creates teacher data including a combination of the attribute data group 221 for each user previously accumulated in the service providing system 5 and the user score 231 determined in advance for the corresponding user (step S101). Then, the machine learning unit 24 inputs the created teacher data into the user score estimation model 230, and generates or updates the user score estimation model 230 used for user score estimation by the user score estimation unit 23 (step S102). Thereafter, the process shown in this flowchart ends. Note that when the attribute estimation unit 22 uses the attribute estimation model 220 for attribute estimation, the attribute estimation model 220 may also be generated and/or updated in a similar process flow.

Figure 7 is a flowchart showing the flow of the user score estimation process according to this embodiment. The process shown in this flowchart is executed for each target user at a timing specified by the administrator.

In steps S201 and S202, the geographic fact attribute 211 is determined. The attribute determination unit 21 estimates the geographic position of the target user based on the behavioral information (behavioral history) of the target user (step S201), and determines the geographic fact attribute 211 related to the geographic position based on the estimated geographic position of the user and predetermined map data (step S202). After that, the process proceeds to step S203.

In steps S203 and S204, the user score 231 is determined and output. The attribute estimation unit 22 estimates the geographical estimated attribute 222 as the attribute data group 221 related to the user based on the determined geographical fact attribute 211 (step S203). In this embodiment, the geographical estimated attribute 222 is estimated as at least a part of the attribute data group 221 as it is, so the attribute estimation unit 22 may obtain the attribute data group 221 of the user by adding the geographical estimated attribute 222 to the attribute data group 221 that is acquired from the service providing system 5 for the target user and is previously stored. Then, the user score estimation unit 23 inputs the attribute data group 221 including the geographical estimated attribute 222 of the target user determined in step S203 into the user score estimation model 230, and acquires the output value as the user score 231 to be set for the user (step S204). However, the estimation method of the user score 231 is not limited to the example in this embodiment. For example, the user score 231 may include a value calculated by inputting the attribute data group 221 into a predetermined function or statistical model that is not a machine learning model. After that, the process shown in this flowchart ends.

The user score 231 set for each user is provided to other systems such as the service providing system 5, and is used to customize services provided to the target user by the other systems such as the service providing system 5.

＜Effects＞
According to this embodiment, it is possible to reflect a context inferred from a geographical location related to a user in the user score 231. That is, according to this embodiment, it is possible to reflect a geographical context that can affect the user score and cannot be obtained from explicit attribute data based on the geographical location related to a user in the user score 231. Furthermore, by using various user attribute data, it is possible to calculate a highly accurate user score 231 even when some attribute data does not exist for a target user.

<First variation>
In the above-described embodiment, an example has been described in which data prepared in advance is used as the map data, but the map data may be newly created for use in the attribute determining unit 21. In this case, the map data may be created, for example, by the following method.

Figure 8 is a diagram showing an outline of the functional configuration of the information processing device 1b according to the first variation. The information processing device 1b functions as an information processing device including an attribute determination unit 21, an attribute estimation unit 22, a user score estimation unit 23, a machine learning unit 24, and a map data creation unit 25, by reading a program recorded in the storage device 14 into the RAM 13 and executing it by the CPU 11, which controls each piece of hardware included in the information processing device 1. Note that in this embodiment and other embodiments described below, each function included in the information processing device 1 is executed by the CPU 11, which is a general-purpose processor, but some or all of these functions may be executed by one or more dedicated processors. In addition, in this variation, the contents of the processing by the attribute determination unit 21, the attribute estimation unit 22, the user score estimation unit 23, and the machine learning unit 24 are roughly similar to those in the embodiment described above, so description thereof will be omitted.

The map data creation unit 25 accepts input of a satellite image and divides the input satellite image into grids of a predetermined size. Here, a "grid" is a satellite image divided using some shape. An example of the shape of the grid is a rectangle, but the shape of the grid is not limited. Other shapes such as a triangle or a hexagon may be adopted as the shape of the grid. The map data creation unit 25 then performs image analysis of the satellite image to estimate what geographical fact attribute 211 each area included in each grid corresponds to, and creates map data by associating the geographical fact attribute 211 estimated for the area corresponding to the grid with the identification information of each grid. Here, examples of the geographical fact attribute 211 estimated by the analysis of the satellite image include urban areas, residential areas, rice fields, mountains, rivers, hills, slopes, curved surfaces, vegetation ranges, etc. As the technology used for image analysis may use AI or machine learning models proposed in the past, details will be omitted, but it is possible to estimate the geographical fact attribute 211 as exemplified above based on color information in the image and shapes and features detected by image analysis.

The map data creation unit 25 also accepts input of a prepared evaluation map (e.g., a risk evaluation map such as a hazard map) and divides the input evaluation map into grids of a predetermined size. The map data creation unit 25 then analyzes the evaluation map to estimate what geographical fact attribute 211 each area included in each grid corresponds to, and creates map data by associating the identification information of each grid with the geographical fact attribute 211 estimated for the area corresponding to the grid. Here, examples of the geographical fact attribute 211 estimated by analyzing the evaluation map include high flood risk and high landslide risk. Note that in this embodiment, a hazard map is given as an example of an evaluation map, but the type of evaluation map that can be adopted is not limited to a hazard map. As the evaluation map, various types of evaluation maps can be used as long as they show an evaluation based on some criteria for points or areas on the map.

By adopting such a creation method, it becomes possible to automate all or part of the map data creation process. Then, by creating and using map data in which geographical fact attributes 211 are attached to each grid, the attribute determination unit 21 can identify the grid related to the user's geographical location, and obtain the geographical fact attributes 211 associated with that grid from the map data, thereby determining the geographical fact attributes 211 related to that user.

As described in the above embodiment, the created map data may be used when the attribute determination unit 21 estimates a geographical position related to a user based on the user's behavioral information (behavioral history) to determine whether the geographical position corresponds to the user's place of residence, place of work, place of study, etc. That is, when the geographical fact attribute 211 of the grid to which the geographical position related to the user corresponds is a residential area, the attribute determination unit 21 can determine that the geographical position indicates the user's place of residence, and when the geographical fact attribute 211 of the grid to which the geographical position related to the user corresponds is an urban area, the attribute determination unit 21 can determine that the geographical position indicates the user's place of work, etc.

<2nd variation>
In the embodiment described above, an example has been described in which the user score estimation unit 23 estimates the user score 231 based on the attribute data group 221 including the geographic estimation attribute 222, the demographic attribute 223, the behavioral attribute 224, and the psychographic attribute 225, but the attribute data group 221 for the user score estimation process may have the following configuration: Note that a description of the configuration that is roughly the same as that of the embodiment described above will be omitted.

9 is a simplified diagram of the user score estimation process according to the second variation. In this variation, the attribute data group 221 may include a geographic fact attribute 211 instead of the geographic estimation attribute 222. In addition to the geographic fact attribute 211 corresponding to the geographic location of the user, the attribute data group 221 may include a geographic fact attribute 211 corresponding to other geographic locations adjacent to the geographic location, a geographic fact attribute 211 corresponding to other geographic locations within a predetermined distance from the geographic location, or a geographic fact attribute 211 corresponding to other geographic locations that can be reached from the geographic location by some means of transportation within a predetermined time. Here, the user score estimation model 230 is trained (learned) using teacher data constructed appropriately according to the type of data constituting the attribute data group 221.

<Third variation>
The attribute data group 221 for the user score estimation process may have the following configuration: Note that a description of the configuration that is roughly the same as the embodiment described above will be omitted.

10 is a simplified diagram of the user score estimation process according to the third variation. In this variation, the attribute data group 221 may include a geographic fact attribute 211 in addition to the geographic estimation attribute 222. Here, similar to the second variation, the attribute data group 221 may include, in addition to the geographic fact attribute 211 and the geographic estimation attribute 222 corresponding to the geographic location of the user, a geographic fact attribute 211 and/or a geographic estimation attribute 222 corresponding to other geographic locations adjacent to the geographic location, a geographic fact attribute 211 and/or a geographic estimation attribute 222 corresponding to other geographic locations within a predetermined distance from the geographic location, and a geographic fact attribute 211 and/or a geographic estimation attribute 222 corresponding to other geographic locations that can be reached from the geographic location within a predetermined time by some means of transportation. Here, the user score estimation model 230 is trained (learned) using teacher data appropriately constructed according to the type of data constituting the attribute data group 221.

<Fourth variation>
The user score estimation unit 23 may estimate (calculate) the user score 231 in the following manner. Note that a description of the configuration that is generally similar to the embodiment described above will be omitted.

FIG. 11 is a simplified diagram of the user score estimation process according to the fourth variation. The user score estimation unit 23 selects and determines one user score estimation model 230 from a plurality of user score estimation models (first user score estimation model 230A, second user score estimation model 230B, third user score estimation model 230C, ... n-th user score estimation model 230N (n is an integer)) according to the geographical estimation attribute 222, and estimates a user score 231 (user score 231A, user score 231B, user score 231C, ... user score 231N) based on the attribute data group 221. Here, n may be > 1, n may be 2, or n may be 3, and there is no limit to the number of user score estimation models 230 indicated by n. In addition, each user score estimation model 230 is trained with teacher data in which at least a part of the geographical estimation attribute 222 is common or similar. When the geographical estimation attribute 222 of the user to be the subject of score estimation indicates a predetermined geographical estimation attribute 222, the user score 231 is estimated using a user score estimation model 230 that has been trained in advance according to the predetermined geographical estimation attribute 222. Note that, for example, at least a part of the geographical estimation attributes 222 being common or similar may mean that the geographical score 222A indicates a predetermined range, and the geographical label 222B is common. Note that in this variation, the attribute data group 221 may further include a geographical fact attribute 211 and/or a geographical estimation attribute 222 related to the geographic location where the user to be the subject of score estimation is located, in addition to the demographic attribute 223, the behavioral attribute 224, the psychographic attribute 225, etc.

<Fifth variation>
The user score estimation unit 23 may estimate (calculate) the user score 231 in the following manner. Note that a description of the configuration that is generally similar to the embodiment described above will be omitted.

Figure 12 is a simplified diagram of the user score estimation process according to the fifth variation. The user score estimation unit 23 may determine the user score 231D by adjusting, correcting, amending, or otherwise modifying the user score 231 obtained by inputting the attribute data group 221 into the user score estimation model 230 based on the geographical estimation attribute 222 relating to the geographical location of the user to be estimated. Here, when the geographical estimation attribute 222 indicates a geographical score 222A that exceeds a predetermined threshold, the user score estimation unit 23 may upwardly correct the user score 231 output by the user score estimation model 230 to determine the user score 231D. On the other hand, when the geographical estimation attribute 222 indicates a geographical score 222A that is below a predetermined threshold, the user score estimation unit 23 may downwardly correct the user score 231 output by the user score estimation model 230 to determine the user score 231D. When correcting the user score 231 output by the user score estimation model 230, the coefficient to be applied to the output user score 231, the score to be added, etc. may be appropriately determined based on the difference between the geographical score 222A and a predetermined threshold value, etc. The user score estimation unit 23 may correct the user score 231 output by the user score estimation model 230 according to the geographical fact attribute 211 instead of the geographical estimation attribute 222, and determine the user score 231D. In this case, the user score estimation unit 23 may correct the user score 231 output by the user score estimation model 230 according to whether or not it is a predetermined geographical fact attribute 211, and determine the user score 231D.

The user score estimation unit 23 may determine that the user score 231 output by the user score estimation model 230 is invalid when the difference between the geographic score 222A in the geographic estimation attribute 222 and the predetermined threshold exceeds another predetermined threshold. Here, the user score estimation unit 23 may determine that the user score 231 output by the user score estimation model 230 is invalid when the amount by which the geographic score 222A in the geographic estimation attribute 222 exceeds the predetermined threshold is a positive value that exceeds another predetermined threshold. Also, here, the user score estimation unit 23 may correct the user score 231 output by the user score estimation model 230 when the amount by which the geographic score 222A in the geographic estimation attribute 222 exceeds the predetermined threshold is a negative value that exceeds another predetermined threshold. Note that the predetermined threshold, the other predetermined threshold, etc. may be set in advance according to the number of user scores 231, and may be set based on a statistical index of the geographic score 222A for each user to whom the user score 231 within a predetermined range is output.

<Other variations>
In the above-described embodiment, an example in which the geographical fact attribute 211 of the geographical position related to the user or the area including the geographical position is acquired has been described, but the geographical fact attribute 211 determined by the attribute determination unit 21 is not limited to such an aspect. For example, the attribute determination unit 21 may acquire the geographical fact attribute 211 for a position or area different from the geographical position related to the user or the area including the geographical position. For example, the attribute determination unit 21 may determine the geographical fact attribute 211 related to the other geographical position based on other geographical positions within a predetermined range based on the geographical position related to the user and predetermined map data. In this case, the attribute estimation unit 22 estimates the geographical estimated attribute 222 as at least a part of the attribute data group 221 related to the user based on the geographical fact attribute 211 related to the determined other geographical position. In addition, for example, the attribute determination unit 21 may determine the geographical fact attribute 211 related to the geographical position of the user according to the geographical fact attribute 211 of another area adjacent to the area including the user's residence, etc. In this way, the attribute estimation unit 22 can estimate the user's attribute data group 221 based not only on the user's place of residence, place of work, etc., but also on the geographical fact attributes 211 of the areas adjacent to these locations.

1. Information processing device

Claims (13)

an attribute determining means for determining a geographic fact attribute associated with a geographic location based on a geographic location associated with a user and predetermined map data;
an attribute estimation means for estimating at least a part of an attribute data group related to the user based on the determined geographic fact attribute;
a user score estimation means for estimating a user score to be set for the user based on the estimated attribute data group,
The map data is data that can identify a past state of a geographical location as the geographical fact attribute corresponding to the geographical location,
The attribute determining means determines the geographic fact attribute related to the geographic location based on a past state corresponding to the geographic location related to the user.
Information processing system. The attribute estimation means estimates a geographical estimated attribute as at least a part of the attribute data group based on the determined geographical fact attribute.
The information processing system according to claim 1 . The attribute estimation means estimates at least a part of the attribute data group based on an output value obtained by inputting the determined geographical fact attribute into an attribute estimation model.
3. The information processing system according to claim 1 or 2 . The attribute estimation means estimates at least a part of the attribute data group by using the attribute estimation model generated using training data in which the determined geographic fact attribute is an input value and a geographic score or a geographic label related to the geographic location is an output value.
The information processing system according to claim 3 . The attribute determination means estimates a geographical position related to the user based on the behavioral information of the user, and determines a geographical fact attribute related to the geographical position based on the estimated geographical position and predetermined map data.
The information processing system according to claim 1 . The attribute determination means determines the geographic fact attribute related to the other geographic location based on other geographic locations within a predetermined range based on the geographic location related to the user and predetermined map data;
The attribute estimation means estimates at least a part of the attribute data group based on the geographical fact attributes related to the determined other geographical locations.
The information processing system according to claim 1 . a map data creation means for creating the map data by dividing a satellite image, estimating a geographical fact attribute corresponding to each divided area by performing image analysis on the satellite image, and associating the geographical fact attribute estimated for each area with identification information of the area;
The attribute determination means determines the geographic fact attribute related to the geographic location based on the map data created by the map data creation means.
The information processing system according to claim 1 . The present invention further comprises a map data creation means for dividing an evaluation map showing an evaluation based on some criteria for points or regions on a map, estimating geographical fact attributes corresponding to each divided region by analyzing the evaluation map, and creating the map data by associating the geographical fact attributes estimated for each region with identification information of the region,
The attribute determination means determines the geographic fact attribute related to the geographic location based on the map data created by the map data creation means.
The information processing system according to claim 1 . The user score estimation means estimates a user score to be set for the user by inputting the attribute data group into a user score estimation model.
The information processing system according to any one of claims 1 to 8 . The user score estimation means estimates the user score using the user score estimation model generated using a machine learning framework based on a gradient boosting decision tree.
The information processing system according to claim 9 . the user score estimation means estimates a user score to be set for the user by using the user score estimation model generated using teacher data having the attribute data group as an input value and the user score related to the user as an output value;
11. The information processing system according to claim 9 or 10 . The computer
an attribute determination step for determining a geographic fact attribute associated with the geographic location based on the geographic location associated with the user and predetermined map data;
an attribute estimation step of estimating at least a portion of an attribute data set related to the user based on the determined geographic fact attributes;
a user score estimating step of estimating a user score to be set for the user based on the estimated attribute data group ;
The map data is data that can identify a past state of a geographical location as the geographical fact attribute corresponding to the geographical location,
The attribute determination step determines the geographic fact attribute associated with the geographic location based on a past state corresponding to the geographic location associated with the user.
method. Computer,
an attribute determining means for determining a geographic fact attribute associated with a geographic location based on a geographic location associated with a user and predetermined map data;
an attribute estimation means for estimating at least a part of an attribute data group related to the user based on the determined geographic fact attribute;
a user score estimation means for estimating a user score to be set for the user based on the estimated attribute data group;
The map data is data that can identify a past state of a geographical location as the geographical fact attribute corresponding to the geographical location,
The attribute determining means determines the geographic fact attribute related to the geographic location based on a past state corresponding to the geographic location related to the user.
program.

Images