JP2021149845A - Damage situation estimation device and computer program - Google Patents

Abstract

To provide a damage situation estimation device capable of reducing a load for investigation of damage situation for insurance money calculation.SOLUTION: A damage situation estimation device is configured to: store policyholder information regarding a policyholder of a disaster insurance that compensates for a damage by a disaster, associating with an actual damage situation by the disaster to the policyholder; refer to the stored policyholder information regarding the policyholder and the actual damage situation according to the policyholder information on a damage situation estimation target policyholder; and estimate a damage situation regarding the disaster of the damage situation estimation target policyholder.SELECTED DRAWING: Figure 2

Description

This technology relates to a damage situation estimation device and a computer program.

Nowadays, various uncertainties about the future have been pointed out, such as the increase in large-scale natural disasters and the uncertainty about the future of the lifetime employment system and social security system. Against this background, there are many cases in which people subscribe to various insurance services such as disaster insurance, life insurance, and automobile insurance in preparation for future risks.

In these insurance services, when damage occurs due to a disaster, accident, etc. and the cause of the insurance claim occurs, the claimant reports the damage to the insurance company. The insurance company conducts an assessment process such as investigation and confirmation, determines whether or not the case is subject to insurance claim payment in the assessment process, and if it is a target case, processes the insurance claim payment. As described above, while various processing is required and it takes time to pay the insurance money, it is desired that the claimant who suffers the damage can pay the insurance money promptly.

As a prior art related to such an insurance business, Patent Document 1 discloses an insurance claim amount validity determination support device that supports determination of the validity of an insurance claim amount due to an insurance accident.

Patent No. 600205

However, in the event of a large-scale disaster, many insurance claims will be made at the same time, so insurance companies need to conduct many investigations in a short period of time. Depending on the insurance content, it was often necessary to check the damage status of the insured object locally, which placed a heavy burden on the person in charge of the insurance company. In addition, in the case of a field survey, it was sometimes necessary for the claimant to be present, and the burden on the claimant was heavy.

An object of the present disclosure is to provide a damage situation estimation device or the like that can reduce the burden of investigating the damage situation for calculating insurance claims.

The damage status estimation device according to one aspect of the present disclosure stores the contractor information regarding the contractor of the disaster insurance that compensates for the damage caused by the disaster and the actual damage status due to the disaster related to the contractor in association with each other, and stores the damage status. Based on the contractor information regarding the estimated target contractor, the damage status related to the disaster of the damage status estimation target contractor is estimated by referring to the stored contractor information regarding the contractor and the actual damage status.

The computer program according to one aspect of the present disclosure stores the policyholder information regarding the contractor of the disaster insurance that compensates for the damage caused by the disaster in association with the actual damage situation caused by the disaster related to the policyholder, and is the target of estimating the damage situation. Based on the contractor information about the contractor, the computer is made to execute the process of estimating the damage situation related to the disaster of the contractor subject to the damage situation estimation by referring to the contractor information about the contractor and the actual damage situation stored. ..

The damage status estimation method according to one aspect of the present disclosure stores the contractor information regarding the contractor of the disaster insurance that compensates for the damage caused by the disaster in association with the actual damage status due to the disaster related to the contractor, and stores the damage status. Based on the contractor information regarding the estimated target contractor, the damage status related to the disaster of the damage status estimation target contractor is estimated by referring to the stored contractor information regarding the contractor and the actual damage status.

According to the present disclosure, it is possible to reduce the burden of investigating the damage situation for calculating insurance claims.

It is a schematic diagram of the information processing system in 1st Embodiment. It is a block diagram which shows the structure of the damage situation estimation device. It is explanatory drawing which illustrates the record layout of the contractor information DB. It is explanatory drawing which illustrates the record layout of the damage situation DB. It is explanatory drawing which illustrates the record layout of the insurance money information DB. It is a block diagram which shows the structure of a terminal device. It is a conceptual diagram which shows the generation method of the estimation model which concerns on 1st Embodiment. It is a conceptual diagram which shows the estimation method of the damage situation using the estimation model which concerns on 1st Embodiment. It is a conceptual diagram which shows the calculation method of the insurance amount which concerns on 1st Embodiment. It is a flowchart which shows an example of the generation processing procedure of the estimation model by the damage situation estimation apparatus. It is a flowchart which shows an example of the estimation processing procedure by a damage situation estimation apparatus. It is a figure which shows an example of the screen displayed on the display part of a terminal device. It is explanatory drawing explaining the structure of the estimation model in 2nd Embodiment.

The present invention will be specifically described with reference to the drawings showing the embodiments thereof. In the following, an example will be described when a flood occurs as a disaster.

(First Embodiment)
FIG. 1 is a schematic diagram of the information processing system 100 according to the first embodiment. The information processing system 100 includes a damage situation estimation device 1, a terminal device 2, and an external server 3. The damage situation estimation device 1 can send and receive information between the terminal device 2 and the external server 3 via a network N such as the Internet.

The damage situation estimation device 1 is an information processing device used by an administrator or the like of an insurance company that provides insurance services. The damage situation estimation device 1 estimates the damage situation of other policyholders based on the actual damage situation of the policyholder of the insurance service. Further, the damage situation estimation device 1 provides information on the insurance service according to the estimation result via the terminal device 2. The terminal device 2 is an information processing device used by a person in charge of an insurance company.

The external server 3 is, for example, a server device managed by a local government, and includes a storage device that stores various external information related to a disaster. The external information includes, for example, damage prediction map information (hazard map information) that predicts the degree of damage that occurs in the event of a disaster in relation to a geographical location, geographical information that indicates geographical characteristics, and the like.

In the present embodiment, the insurance company mainly provides disaster insurance services for large-scale natural disasters where widespread damage is expected, and the insurance target is the real estate (house) owned by the policyholder. do. Examples of large-scale natural disasters include flood disasters, snow disasters, and earthquake disasters. The insurance company provides a disaster insurance service to the policyholders of the disaster insurance, which pays the insurance money according to the degree of damage caused by the disaster based on the contract contents. In the event of a disaster, the person in charge of the insurance company will conduct a field survey at the contractor's house that was damaged by the disaster and obtain the actual damage status of the house. The person in charge transmits the acquired actual damage situation to the damage situation estimation device 1 using the terminal device 2. The damage situation estimation device 1 collects the damage situation from a plurality of field surveys, and creates an estimation model of the damage situation for the disaster based on the collected damage situation. When an insurance claim is received from a new contractor (contractor subject to damage status estimation), the damage status estimation device 1 calculates an estimated value of the damage status using an estimation model.

The configuration and detailed processing contents of such an information processing system 100 will be described below.

FIG. 2 is a block diagram showing the configuration of the damage situation estimation device 1. The damage situation estimation device 1 is, for example, a server computer. The damage situation estimation device 1 includes a control unit 10, a storage unit 11, and a communication unit 12. The damage situation estimation device 1 is described as one server computer for ease of explanation, but the functions or processes may be distributed among a plurality of server computers, and the damage situation estimation device 1 is virtually generated by one large computer. It may be one of a plurality of server computers (instances).

The control unit 10 is a processor using one or a plurality of CPUs (Central Processing Units), GPUs (Graphics Processing Units), etc., and uses a built-in memory such as a ROM (Read Only Memory) or a RAM (Random Access Memory). , Control each component to execute processing.

The communication unit 12 is a communication interface that realizes communication via the network N. The control unit 10 can send and receive information to and from the terminal device 2 via the network N by the communication unit 12.

The storage unit 11 includes a non-volatile storage medium such as a hard disk or an SSD (Solid State Drive). The storage unit 11 stores programs and data referred to by the control unit 10 including the program 1P. The control unit 10 reads and executes the program 1P to make a general-purpose server computer function as a damage situation estimation device peculiar to the present disclosure.

The program 1P stored in the storage unit 11 may be recorded on a recording medium so that it can be read by a computer. The storage unit 11 stores the program 1P read from the recording medium 1A by a reading device (not shown). Further, the program 1P may be downloaded from an external computer (not shown) connected to a communication network (not shown) and stored in the storage unit 11. The storage unit 11 may be composed of a plurality of storage devices, or may be an external storage device connected to the damage situation estimation device 1.

Further, the storage unit 11 stores the contractor information DB (Data Base: database) 111, the damage status DB 112, the insurance money information DB 113, the estimation model 1M, and the like. The estimation model 1M is a calculation model for estimating the damage situation. The estimation model 1M will be described later.

FIG. 3 is an explanatory diagram illustrating the record layout of the contractor information DB 111. The policyholder information DB 111 is a database that stores information about policyholders of insurance services. The damage situation estimation device 1 acquires information about the policyholder at the time of contracting the insurance service and stores it in the policyholder information DB 111. The contractor information DB 111 stores, for example, the attribute information of the contractor, the living environment information of the contractor, the contract plan, and the like in association with the contractor ID of each of the plurality of contractors.

The contractor ID is identification information that identifies the contractor. The attribute information includes, for example, the contractor's name. The living environment information is information on the housing environment of the contractor, and includes items such as the structure of the house, the number of floors, the age of the house, and the address. The structure includes information indicating the building material of the house, and stores, for example, wooden structure, reinforced concrete, and the like. The structure may include a detached house, a condominium, etc. relating to the structure of a house. The number of floors includes information indicating the number of floors of a house or the floor on which a house exists, and stores, for example, two floors, two floors, and the like. The age of the house includes information indicating the age of the house since it was built. The address includes information that identifies the location (area) of the house, for example, an address address is stored. Latitude / longitude information may be stored in the address. The contract plan is information that identifies the insurance content of the policyholder, and includes, for example, the name of the contract plan. The stored contents of the contractor information DB 111 are not limited to the example shown in FIG.

FIG. 4 is an explanatory diagram illustrating the record layout of the damage status DB 112. The damage status DB 112 is a database that stores information on the damage caused by a disaster. For example, the damage status DB 112 stores the actual damage status of the contractor in association with the contractor ID of the contractor who has suffered damage. The actual damage situation is information on the actual damage situation to the contractor's house. The actual damage situation includes, for example, the height of flooding in a house, damage to a building, and the like. Damage to a building is classified into, for example, total damage, partial damage, partial damage, etc. according to the degree of damage. These actual damage situations are the actual damage situations identified after the person in charge directly confirmed the damaged house at the site where the contractor's house is located. After the field survey, the damage status estimation device 1 acquires the actual damage status from the person in charge via the terminal device 2 and collects it in the damage status DB 112. The stored contents of the damage status DB 112 are not limited to the example shown in FIG.

FIG. 5 is an explanatory diagram illustrating the record layout of the insurance claim information DB 113. The insurance money information DB 113 is a database that stores information related to the estimation of insurance money. The damage situation estimation device 1 stores information regarding the calculation of the insurance money to be paid in the insurance service contract in advance in the insurance money information DB 113. The insurance money information DB 113 includes calculation point information and insurance amount information. In the calculation point information, for example, the damage status and the calculation point used for calculating the insurance amount are stored in association with each other. The damage status includes multiple items such as flood height and damage to buildings, and calculation points are associated with each item. For example, the contract plan, the calculation point, and the insurance amount are stored in association with the insurance amount information. The insurance amount paid to the policyholder is determined, for example, by the contract plan and the total points calculated. The stored contents of the insurance claim information DB 113 are not limited to the example shown in FIG.

FIG. 6 is a block diagram showing the configuration of the terminal device 2. The terminal device 2 is, for example, a smartphone, a tablet terminal, a personal computer, or the like. The terminal device 2 includes a control unit 20, a storage unit 21, a communication unit 22, a display unit 23, and an operation unit 24.

The control unit 20 includes a processor such as a CPU or GPU, a memory, and the like. The control unit 20 may be configured as one piece of hardware (SoC: System On a Chip) in which a processor, a memory, a storage unit 21, and a communication unit 22 are integrated. The control unit 20 controls each component unit to execute processing based on the program stored in the storage unit 21.

The storage unit 21 includes a non-volatile memory such as a flash memory. The storage unit 21 stores the program and data referred to by the control unit 20. The communication unit 22 is a communication interface that realizes communication with the damage situation estimation device 1 via the network N.

The display unit 23 includes a display device such as a liquid crystal panel and an organic EL (Electro Luminescence) display. The display unit 23 displays various information according to the instruction from the control unit 20. The operation unit 24 is an interface that accepts user operations, and includes, for example, a physical button, a mouse, a touch panel device with a built-in display, a speaker, a microphone, and the like. The operation unit 24 receives an operation input from the user and sends a control signal according to the operation content to the control unit 20.

FIG. 7 is a conceptual diagram showing a method of generating the estimation model 1M according to the first embodiment. The method of generating the estimation model 1M will be specifically described with reference to FIG. 7.

The damage situation estimation device 1 generates an estimation model 1M using the contractor information DB 111, the damage situation DB 112, and external information. The damage status estimation device 1 acquires the living environment information associated with the contractor ID from the contractor information DB 111 and the actual damage status associated with the contractor ID from the damage status DB 112. Further, the damage situation estimation device 1 acquires hazard map information as external information from the external server 3 and uses it as additional information of living environment information. In the example of FIG. 7, the hazard map information includes an address including the residential area of the contractor and a water depth rank that predicts inundation damage to the address in stages. The damage status estimation device 1 is based on the acquired living environment information, actual damage status, and hazard map information, and has a hazard map according to the living environment information and damage status related to the same contractor ID and the address included in the living environment information. Generate damage status log data associated with information. When a predetermined period or a predetermined number of damage status log data is collected, the damage status estimation device 1 generates an estimation model 1M that estimates the damage status for each living environment based on the collected log data.

The damage status estimation device 1 separates the collected damage status log data for each living environment. The living environment is determined based on the living environment information and hazard map information. For example, the damage status log data in which all the values of each item match based on the matching rate of the values of each item of the structure, the number of floors, the age of the building, the address, and the water depth rank of the hazard map information included in the living environment information is Sorted into the same living environment group. For items such as address and age, numerical values indicating a predetermined range such as 1-chome 1 to 6 and 6 to 10 years may be set. Further, the items in which all the values of each item match are not limited to the same group, and the items in which the matching rate in a plurality of items is equal to or more than a predetermined value may be included in the same group.

The damage status estimation device 1 calculates the average value of each item of the actual damage status (for example, inundation height, damage to the building) in all the damage status log data included in the same living environment group. Derive the estimated value of the damage situation for the living environment information. Alternatively, the damage situation estimation device 1 may derive the estimated value of the damage situation for the living environment information by acquiring the mode value of each item of the damage situation for the same living environment information. The damage situation estimation device 1 generates an estimation model 1M in which the estimated damage situation, which is the derived estimated value, is associated with the living environment information and the hazard map information. In this way, based on the contractor information, external information, and actual damage information of a plurality of contractors, and the contractor information and external information of the contractor whose damage status is estimated, the contractor information and the external information are used as keys for damage. A estimation model 1M that outputs an estimated value of the damage status of the contractor subject to status estimation is generated.

In the above, an example in which hazard map information is used as external information in generating damage status log data has been described, but the present embodiment is not limited. As the external information, the damage situation estimation device 1 may use geographical information indicating geographical characteristics at a geographical position provided by, for example, a local government. In the geographic information, for example, geographical characteristic information such as land elevation and topography is recorded in association with an address (location information). The geographic information may be topographic map information. The damage status estimation device 1 generates damage status log data in which the living environment information included in the contractor information, the geographical information associated with the address and the geographical characteristic information (for example, the altitude), and the actual damage status are associated with each other. The damage situation estimation device 1 generates an estimation model 1M that derives an estimated value of the actual damage situation using the living environment information and the geographical information as keys based on the generated damage situation log data.

As the external information, the damage situation estimation device 1 may use, for example, damage map information showing the actual damage state provided by the local government or the like after the occurrence of a disaster in association with the location information. The damage status estimation device 1 is a damage status log in which the living environment information included in the contractor information, the damage map information associated with the address and the actual damage status (for example, the actual water depth rank), and the actual damage status are associated with each other. Generate data. The damage situation estimation device 1 generates an estimation model 1M that derives an estimated value of the actual damage situation using the living environment information and the damage map information as keys based on the generated damage situation log data.

The damage situation estimation device 1 may use weather information, rain cloud radar information, and the like as external information. Alternatively, the damage status estimation device 1 may generate damage status log data based only on the living environment information and the damage status without adding external information. Further, the damage status log data used for generating the estimation model 1M is not limited to the actual damage status for the same disaster, and for example, the damage status data collected in the disasters that occurred in the past may be used.

FIG. 8 is a conceptual diagram showing a method of estimating the damage situation using the estimation model 1M according to the first embodiment. The method of estimating the damage situation will be specifically described with reference to FIG.

If the person in charge receives an insurance claim from a policyholder (hereinafter referred to as a policyholder subject to damage status estimation) for an insurance contract for which the damage status has not been specified after the estimation model 1M is generated, the person in charge will be damaged locally. Instead of confirming the situation, communication with the damage situation estimation device 1 is performed via the terminal device 2 to acquire the damage situation estimation result.

The damage situation estimation device 1 acquires an estimated value of the damage situation using the contractor information DB 111, the hazard map information, and the estimation model 1M. The damage status estimation device 1 acquires contractor information including living environment information of the contractor subject to damage status estimation from the contractor information DB 111 based on the contractor ID of the contractor subject to damage status estimation. Further, the damage status estimation device 1 acquires hazard map information corresponding to the address of the contractor subject to damage status estimation from the external server 3.

The damage situation estimation device 1 uses the estimation model 1M to derive an estimated value (estimated damage situation) of the damage situation according to the living environment of the contractor whose damage situation is estimated. In this case, the damage situation estimation device 1 sets the matching rate between each item of the living environment information and the hazard map information in the estimation model 1M and each item of the living environment information and the hazard map information of the contractor subject to the damage situation estimation, respectively. By calculating, an estimated value of the damage situation may be derived. The damage status estimation device 1 identifies the living environment with the highest total value of the matching rate for each item, and uses the damage status corresponding to the specified living environment as the estimated value of the damage status to the living environment of the contractor subject to damage status estimation. Derived. In calculating the match rate, different weighting may be set for each item. Further, in specifying the living environment, conditions may be set such that only items having a matching rate of a predetermined value or more are selected for a predetermined item. The damage situation estimation device 1 outputs the derived estimated value of the damage situation to the terminal device 2. The person in charge inputs the estimated value of the damage situation as the damage situation of the contractor subject to the damage situation estimation, for example, in the survey form.

In this way, by using the estimation model 1M generated based on the actual damage situation actually obtained from the contractor who entered the damage situation in the past with a similar living environment, the person in charge can contact all the contractors locally. It is possible to specify the estimated information of the damage situation while reducing the burden of directly visually recognizing the house.

FIG. 9 is a conceptual diagram showing a method of calculating the insurance amount according to the first embodiment. The method of calculating the insurance amount will be specifically described with reference to FIG.

The damage situation estimation device 1 acquires the estimated amount of insurance amount of the damage situation estimation target contractor by using the contractor information of the damage situation estimation target contractor, the damage situation estimation value, and the insurance money information DB 113. The damage situation estimation device 1 reads out the survey form data or the information stored in the database, and acquires the contractor information including the contract plan of the contractor subject to the damage situation estimation and the estimated value of the damage situation. The damage situation estimation device 1 refers to the calculation point information of the insurance money information DB 113 and acquires the calculation points corresponding to the contents of each item of the estimated damage situation. In the example of FIG. 9, since the estimated value of the damage situation of the contractor subject to the damage situation estimation matches the item inundation height of 1.0 m or more and less than 1.5 m, the calculation point for the item inundation height is 20 points. .. Similarly, since the damage status of the contractor subject to damage status estimation coincides with the partial destruction in the damage of the item building, the calculation point for the damage of the item building is 20 points. Therefore, the total value of the calculated points with respect to the estimated value of the damage situation of the contractor subject to the damage situation estimation is 40 points.

The damage situation estimation device 1 refers to the insurance money information of the insurance money information DB 113 and specifies the insurance amount according to the total value of the contract plan and the calculation points. In the example of FIG. 9, the insurance amount corresponding to the total calculation points 31 to 40 points of Plan C is 10 million yen. In this way, the damage situation estimation device 1 calculates the estimated amount of insurance amount based on the estimated value of the damage situation. The damage situation estimation device 1 transmits the calculation result to the terminal device 2.

FIG. 10 is a flowchart showing an example of a procedure for generating an estimation model 1M by the damage situation estimation device 1. The following processing is executed by the control unit 10 according to the program 1P stored in the storage unit 11 of the damage situation estimation device 1.

The control unit 10 of the damage situation estimation device 1 acquires contractor information including the contractor's contractor ID, attribute information, living environment information, contract plan, etc., for example, when the contractor subscribes to the insurance service (step S11). The control unit 10 stores the acquired contractor information in the contractor information DB 111. Further, the control unit 10 acquires hazard map information as external information from the external server 3 of the local government or the like (step S12), and stores it in the storage unit 11.

In the event of a disaster and damage to the policyholder's home, the insurance company representative will conduct a field survey of the policyholder's home. The person in charge confirms the inundation height of the house, the degree of damage to the building, etc. in the presence of the contractor, and identifies the actual damage situation. The person in charge creates a survey form by inputting the contractor ID, the actual damage situation, etc. using the operation unit 24 or the like of the terminal device 2. The control unit 20 of the terminal device 2 acquires survey form data including the contractor ID, the actual damage situation, and the like. The control unit 20 associates the contractor ID with the acquired actual damage status and transmits it to the damage status estimation device 1.

The control unit 10 of the damage status estimation device 1 acquires the actual damage status associated with the contractor ID by the communication unit 12 (step S13). The control unit 10 stores the acquired actual damage status and the contractor ID in the damage status DB 112 (step S14). As a result, the actual damage status and the contractor information are associated with each other based on the contractor ID and stored in the storage unit 11.

The control unit 10 collects the damage status for the same disaster by repeatedly executing the above-mentioned processing for a plurality of contractors. The control unit 10 generates damage status log data based on the collected damage status, living environment information of contractor information, and hazard map information (step S15). When a predetermined period or a predetermined number of damage status log data is generated, the control unit 10 generates an estimation model 1M that outputs an estimated value of the damage status for each living environment by analyzing the damage status log data.

Specifically, the control unit 10 sorts the damage status log data for each living environment based on the matching rate of each item of the living environment information and the hazard map information included in each collected damage status log data (step S16). ). The control unit 10 derives an estimated value of the damage status for the living environment information by calculating the average value of each item of the actual damage status in all the damage status log data included in the same living environment group. (Step S17), the estimation model 1M is generated (step S18).

Through the above processing, the damage situation estimation device 1 generates an estimation model 1M for estimating the damage situation of a house in a similar living environment based on the living environment information of the damaged house at the time of a disaster and the actual damage situation. Since the estimation model 1M is generated based on a large number of actual damage situations for the disaster, it is possible to output estimation information according to the actual situation of the disaster. Furthermore, by adding external information such as hazard map information, the damage situation can be estimated based on the damage situation log similar to the living environment even in the area where the actual damage situation is not acquired.

Using the estimation model 1M generated as described above, the damage situation estimation device 1 executes the estimation process. FIG. 11 is a flowchart showing an example of the estimation processing procedure by the damage situation estimation device 1. The following processing is executed by the control unit 10 according to the program 1P stored in the storage unit 11 of the damage situation estimation device 1.

After the estimation model 1M is generated, the person in charge of the insurance company accepts a new insurance claim from the policyholder for which the damage situation is estimated. The person in charge uses the terminal device 2 to specify a contractor ID or the like that identifies the contractor subject to damage status estimation, and inquires about the estimated value of the damage status related to the contractor subject to damage status estimation. Send to.

The control unit 10 of the damage status estimation device 1 receives the inquiry, refers to the contractor information DB 111, and contractor information related to the contractor ID of the received inquiry request (contractor information regarding the contractor subject to damage status estimation). (Step S21).

The control unit 10 uses the estimation model 1M to estimate the damage status according to the contractor information of the contractor for which the damage status is estimated. Specifically, the control unit 10 has, for example, a living environment including living environment information included in the contractor information of the contractor subject to damage status estimation and hazard map information corresponding to the address of the contractor subject to damage status estimation, and an estimation model. Based on the living environment included in 1M, the matching rate between each item of the living environment related to the contractor subject to damage situation estimation and each item of the living environment included in the estimation model 1M is calculated (step S22). The control unit 10 identifies a living environment group similar to the living environment of the contractor whose damage status is estimated based on the total value of the calculated matching rates (step S23). The control unit 10 acquires the damage status associated with the specified living environment group as an estimated value of the damage status of the contractor whose damage status is estimated (step S24).

The control unit 10 automatically inputs the acquired estimated value of the damage situation into the survey form data of the contractor subject to the damage situation estimation. The control unit 10 generates screen information including an estimated value of the damage situation that has been automatically input. The control unit 10 outputs screen information including the generated estimated value of the damage situation to the terminal device 2 (step S25). The person in charge recognizes the estimated value of the damage situation of the contractor for which the damage situation is estimated by the display unit 23 of the terminal device 2.

Further, the control unit 10 calculates the estimated amount of insurance amount of the policyholder subject to the damage situation estimation based on the survey form data in which the estimated value of the damage situation is input and the insurance money information DB 113 (step S26). The control unit 10 generates screen information including an estimated insurance amount based on the calculation result. The control unit 10 outputs screen information including the estimated amount of the generated insurance amount to the terminal device 2 (step S27), and ends a series of processes. The person in charge recognizes the estimated amount of insurance amount of the contractor subject to damage situation estimation by the display unit 23 of the terminal device 2.

In the above, an example in which the estimation model 1M is generated in advance and the damage situation is estimated using the generated estimation model 1M has been described, but the present embodiment is not limited. The estimation model 1M may be generated in response to an inquiry about the estimated value of the damage status of the contractor whose damage status is estimated. Further, the estimation model 1M may be generated each time a new estimation value is inquired.

In the processes described in the flowcharts of FIGS. 10 and 11 described above, a part or all of the processes executed by the control unit 10 of the damage situation estimation device 1 are externally communicably connected to the damage situation estimation device 1. It may be executed by using the device, or may be executed by the control unit 20 of the terminal device 2.

FIG. 12 is a diagram showing an example of a screen displayed on the display unit 23 of the terminal device 2. The control unit 20 of the terminal device 2 displays the screen shown in FIG. 12 on the display unit 23 based on the screen information received from the damage situation estimation device 1. FIG. 12A shows an example of screen 231 including an estimated damage situation. As shown in FIG. 12A, the screen 231 including the estimated damage status shows the contractor ID including the contractor for which the damage status is estimated, the contractor name, the living environment information, the contract details, and the hazard map. Information and estimated damage status are displayed. When the control unit 10 acquires the estimation result of the damage situation by the estimation model 1M, the control unit 10 acquires the contractor information, hazard map information, and the like of the contractor ID corresponding to the estimation result read from the contractor information DB 111. The control unit 10 generates screen information for displaying the acquired information in association with each other, and transmits the generated screen information to the terminal device 2.

The screen 231 including the estimated value of the damage situation may be in a format corresponding to a predetermined survey form. The control unit 10 outputs the survey form data in which the estimated value of the damage situation is automatically input to the damage status column of the survey form via the terminal device 2. The person in charge confirms the screen contents and inputs a judgment on the estimated value of the damage situation by clicking an operation button such as an OK button or a change button. When the input of the OK button is accepted, the damage situation estimation device 1 calculates the insurance amount based on the estimated value of the damage situation. When the input of the change button is accepted, the damage situation estimation device 1 may update the survey form data based on the newly acquired damage situation after the change, and calculate the insurance amount based on the updated contents. As a result, the burden of inputting the damage status of the person in charge in creating the survey form data is reduced, and the occurrence of erroneous input is prevented.

FIG. 12B shows an example of screen 232 including an estimated amount of insurance. As shown in FIG. 12B, on the screen 232 including the estimated amount of insurance amount, the contractor ID including the contractor, the contractor name, the living environment information, the contractor information including the contract contents, the hazard map information, and the damage are displayed. The estimated value of the situation and the estimated amount of insurance amount are displayed. When the control unit 10 acquires the estimated amount of insurance amount according to the estimated result of the damage situation, the contractor information, the hazard map information, and the damage situation of the contractor ID corresponding to the estimated result read from the contractor information DB 111. Get the estimated value of. The control unit 10 generates screen information for displaying the acquired information in association with each other, and transmits the generated screen information to the terminal device 2.

According to the above processing, the damage situation estimation device 1 estimates the damage situation of the contractor using the estimation model 1M. Furthermore, the estimated amount of insurance is calculated according to the estimation result. As a result, for example, the person in charge of the insurance company can reduce the work load of actually investigating and confirming the damage situation of all the houses when calculating the insurance amount, speeding up the processing in the insurance money service and reducing the labor cost. It is possible to reduce it. Furthermore, since it is not necessary for the contractor to witness a field survey to determine the damage situation, the burden of using the insurance service is reduced.

(Second Embodiment)
In the second embodiment, a configuration using a learning model configured by a neural network as the estimation model 1M will be described. Hereinafter, the differences between the second embodiment and the first embodiment will be described. Since the other configurations other than the configurations described later are the same as those of the information processing system 100 of the first embodiment, the common configurations are designated by the same reference numerals and detailed description thereof will be omitted.

The damage situation estimation device 1 of the second embodiment stores the estimation model 1M, which is a learning model configured by the neural network, in the storage unit 11. The estimation model 1M is a learning model generated by machine learning, and outputs data indicating a damage situation in response to input of living environment information and hazard map information. The estimation model 1M is defined by the definition information. The definition information of the estimation model 1M includes, for example, structural information and layer information of the estimation model 1M, node information included in each layer, and learned parameters. The storage unit 11 stores definition information regarding the estimation model 1M.

FIG. 13 is an explanatory diagram illustrating the configuration of the estimation model 1M in the second embodiment. The estimation model 1M is generated and learned in advance by deep learning using a neural network in the damage situation estimation device 1 or an external device. The learning algorithm may be a recurrent neural network (RNN) when time series data is acquired.

In the example shown in FIG. 13, the estimation model 1M has an input layer for inputting living environment information and hazard map information (external information), an output layer for outputting data indicating a damage situation, and an intermediate layer for extracting feature quantities ( It has a hidden layer). The intermediate layer has a plurality of nodes for extracting the features of the input data, and passes the features extracted using various parameters to the output layer. When living environment information and hazard map information are input to the input layer, calculations are performed in the intermediate layer according to the learned parameters, and output information indicating the damage status is output from the output layer.

The input information input to the input layer of the estimation model 1M is living environment information and hazard map information. Each item information included in the living environment information (for example, structure, number of floors, age, address, etc.) and each item information included in the hazard map information (for example, address, water depth rank, etc.) are input to the input layer. The external information included in the input information is not limited to the hazard map information, and may be other geographic information, damage map information, weather information, rain cloud radar information, and the like.

The estimation model 1M includes a plurality of output layers that output estimation data of items included in the damage situation, respectively. The output information output from the output layer of the estimation model 1M is, for example, data indicating the inundation height and the damage of the building. The inundation height output layer includes nodes corresponding to the set inundation height values (for example, 0 m, 0.1 m, 0.2 m, ...), And the accuracy for each inundation height value is used as a score. Output. The damage situation estimation device 1 can use the value of the inundation height having the highest score or the value of the inundation height having the score equal to or higher than the threshold value as the output value of the output layer. The output layer may have one output node that outputs the most accurate inundation height value instead of having a plurality of output nodes that output the accuracy of each inundation height value.

The output layer of the damage of the building includes the nodes corresponding to the set damage values (for example, total damage, partial damage, partial damage, etc.), and outputs the accuracy for each damage value as a score. The damage situation estimation device 1 can use the damage value having the highest score or the damage value having a score equal to or higher than the threshold value as the output value of the output layer. Note that the output layer may have one output node that outputs the most accurate damage value instead of having a plurality of output nodes that output the accuracy of each damage value.

By collecting the contractor information, the actual damage status and the hazard map information, the control unit 10 assigns the known actual damage status specified by the person in charge by the field survey to the input information including the living environment information and the hazard map information. The obtained information group is acquired as training data. The control unit 10 learns the estimation model 1M using the acquired training data. The control unit 10 learns various parameters and weights constituting the estimation model 1M by using, for example, an error backpropagation method so as to output the damage status according to the living environment information and the hazard map information.

The estimation model 1M is not limited to the example shown in FIG. The estimation model 1M may be a model trained by another algorithm such as a support vector machine that does not use a neural network or a regression tree.

Using the estimation model 1M configured as described above, the damage situation estimation device 1 executes the damage situation estimation process. The damage situation estimation device 1 inputs the living environment information and hazard map information of the contractor information for which the damage situation is estimated to the input layer of the estimation model 1M, and acquires the damage situation estimation data output from the output layer of the estimation model 1M. do.

According to this embodiment, the damage situation can be estimated accurately by using the estimation model 1M composed of the neural network. In the rule base, it is necessary to prescribe all the rules in advance, and it is very complicated to prescribe a complicated condition including a large number of items. However, an estimation result can be easily obtained by using a learning model.

The embodiments disclosed this time are exemplary in all respects and are not restrictive. The scope of the present invention is indicated by the scope of claims and includes all modifications within the meaning and scope equivalent to the scope of claims.

1 Damage status estimation device 2 Terminal device 3 External server 10,20 Control unit 11,21 Storage unit 111 Contractor information DB
112 Damage status DB
113 Insurance claim information DB
1P program 1M estimation model

Claims (6)

The policyholder information about the disaster insurance policyholder who compensates for the damage caused by the disaster is stored in association with the actual damage situation caused by the disaster related to the policyholder.
Based on the contractor information about the contractor subject to damage status estimation, the damage status related to the disaster is estimated by referring to the contractor information about the contractor and the actual damage status stored. Estimator. The damage situation estimation device according to claim 1, wherein the damage situation is estimated by using the damage prediction map information regarding the damage of the disaster. The damage situation estimation device according to claim 1 or 2, which estimates the damage situation using geographic information. The damage situation estimation device according to any one of claims 1 to 3, which estimates the insurance amount related to the disaster of the contractor subject to the damage situation estimation based on the damage situation. The contractor information includes living environment information including at least one of the structure, number of floors, age, and address of the house owned by the contractor.
Claims 1 to 4 for estimating the damage situation according to the living environment information of the contractor information regarding the contractor for which the damage situation is estimated, based on the living environment information of the contractor information regarding the contractor and the actual damage situation. The damage situation estimation device according to any one of the above items. The policyholder information about the disaster insurance policyholder who compensates for the damage caused by the disaster is stored in association with the actual damage situation caused by the disaster related to the policyholder.
Based on the contractor information regarding the contractor subject to damage status estimation, the process of estimating the damage status related to the disaster of the contractor subject to damage status estimation is performed by referring to the contractor information regarding the contractor and the actual damage status stored. A computer program that lets a computer run.

Images