Insurance underwriting automation device using significant disease information judgment and risk classification model

Abstract

The present invention relates to an automated insurance contract review process utilizing a model for determining significant disease information and classifying risk. The present invention comprises: a significant disease information extraction unit that extracts significant disease information of a customer from insurance claim payment information of the customer accumulated in a credit information concentration agency and stores the extracted significant disease information in a health information DB; a mandatory disclosure disease information selection unit that selects mandatory disclosure disease information by applying an insurance company's underwriting criteria information to the significant disease information stored in the health information DB and stores the selected mandatory disclosure disease information in a mandatory disclosure disease information DB; and a personal disease history customized coverage recommendation unit that recommends coverages available for subscription to the customer based on the customer's subscription information, significant disease information stored in the health information DB, and mandatory disclosure disease information stored in the mandatory disclosure disease information DB. According to the present invention, there is an effect of providing an automated insurance contract review process utilizing a significant disease information judgment and risk classification model, which enables insurance policyholders and insurance solicitors to check the conditions for and eligibility of an insurance contract in real time, drastically reduce the number of mandatory disclosure disease items, and allow insurers to operate an efficient insurance contract review organization by simplifying the insurance application process.

Description

Technology Field

[0001] The present invention relates to an automated insurance contract review process utilizing a model for determining significant disease information and classifying risk. Background Technology

[0002] Personal insurance, including life insurance and health insurance, calculates premiums through the development of risk rates for the same risk group, and products are structured to pay insurance benefits for pre-agreed insured events.

[0003] Due to the characteristics of such life insurance, when applying for an insurance contract, a contract review (underwriting) process is required to verify whether the insured is classified into the risk group predicted during the calculation of the risk rate, and to make comprehensive decisions regarding the acceptance of the insurance contract, premiums, and conditions of subscription. In accordance with the Commercial Act, the insurer, the insurance company, must send a notice of acceptance or rejection of the contract through the contract review process to the counterparty within 30 days from the date of receiving all or part of the premium amount along with the application for the insurance contract.

[0004] An application for a life insurance contract is a declaration of intent by the policyholder to enter into a specific insurance contract with an insurer (insurance company), and the act of application is carried out using an application form and the duty to disclose information prior to the contract. Based on the application form and the duty to disclose information prior to the contract prepared by the policyholder and the insured, the underwriter undergoes a contract review process in which they verify all disclosure information listed in the documents, directly check the insured's health status in accordance with each insurance company's disease screening guidelines, and decide whether to accept the contract. Since the underwriter directly verifies all information and documents collected during the life insurance contract application and performs the contract review, it generally takes at least 2 days and up to 30 days from the time the application is submitted until the final acceptance of the contract is confirmed.

[0005] As described above, due to the lengthy life insurance contract review process, there is a problem in that it is difficult for policyholders and insurance agents to quickly determine whether a contract will be accepted.

[0006] If an insurance policyholder misremembers their past medical history and enters incorrect information on the application form or pre-contract disclosure form, or if additional information needs to be provided, the insurance agent must meet with the customer again to supplement the documents. This ultimately prevents the agent from focusing on the sales activities they are essentially required to perform, which can have a direct or indirect negative impact on the insurance company's revenue and lead to increased cost burdens for the company.

[0007] In other words, the employment of contract reviewers, which is necessary for the contract review process that must be performed manually, must be continuously maintained; this ultimately leads to increased business expenses and can act as a factor that worsens the profitability of insurance companies. Prior art literature

[0008] Registered Patent Publication No. 10-2595654 (Registration Date: October 25, 2023, Title: System and Method for Predicting the Probability of Passing Underwriting of Insurance Contracts Using Digital Medical Record Information) Registered Patent Publication No. 10-2646104 (Registration Date: March 6, 2024, Title: AI Assistant System Equipped with Online Pre-Underwriting Function and Method Thereof) Registered Patent Publication No. 10-2394888 (Registration Date: May 2, 2022, Title: Method for Preventing Omission of Medical History Subject to Duty of Disclosure) The problem to be solved

[0009] The technical objective of the present invention is to provide an automated insurance contract review process utilizing a significant disease information judgment and risk classification model, which enables insurance policyholders and insurance agents to check the conditions for and eligibility of an insurance contract in real time, drastically reduce the number of items of diseases requiring disclosure (hereinafter, mandatory disclosure diseases), and allow insurers to operate an efficient insurance contract review organization through the simplification of the insurance application process. means of solving the problem

[0010] The present invention, for solving such technical problems, is an automated insurance contract screening process utilizing a significant disease information judgment and risk classification model, comprising: a significant disease information extraction unit that extracts a customer's significant disease information from the customer's insurance claim payment information accumulated at a credit information concentration agency and stores the extracted significant disease information in a health information DB; a mandatory disclosure disease information selection unit that selects mandatory disclosure disease information by applying an insurer's underwriting criteria information to the significant disease information stored in the health information DB and stores the selected mandatory disclosure disease information in a mandatory disclosure disease information DB; and a personal disease history customized coverage recommendation unit that recommends coverages available for subscription to the customer based on the customer's subscription information, the significant disease information stored in the health information DB, and the mandatory disclosure disease information stored in the mandatory disclosure disease information DB.

[0011] In an automated insurance contract review process utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit determines the validity of each treatment content corresponding to each payment information using the disease classification code (KCD), treatment details (surgery, hospitalization, outpatient), treatment start date, treatment end date, date of occurrence of the insured event, date of payment of insurance benefits, number of treatments, number of treatment days, and total treatment period included in the insurance benefit payment information message transmitted from the credit information concentration agency, and determines a reference date to be used for calculating the elapsed period by applying a pre-set reference date determination rule to the treatment content determined to be valid.

[0012] In an automated insurance contract review process device utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit is characterized by using the reference date closest to the current time point for calculating the elapsed period when multiple reference dates exist for the same disease classification code.

[0013] In an automated insurance contract review process utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit assigns a pre-set priority to treatment contents corresponding to each payment information included in the insurance payment information text, and if multiple treatment contents are valid, selects one treatment content according to the priority to generate the health information DB.

[0014] In an insurance contract review process automation device utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit is characterized by generating the health information DB by selecting one treatment content according to the priority when multiple treatment contents exist for the same disease classification code.

[0015] In an automated insurance contract review process utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit is characterized by generating the health information DB by selecting the highest amount for the payment amount, selecting the maximum number of treatments for each treatment content, selecting the maximum number of treatment days, and selecting the maximum period among each payment case for the total treatment period when there are multiple insurance payout cases with the same disease classification code and reference date.

[0016] In an automated insurance contract review process utilizing a significant disease information judgment and risk classification model according to the present invention, the significant disease information extraction unit is characterized by generating the health information DB by summing the respective payment amounts, summing the number of treatments and treatment days by treatment content respectively, and summing the treatment periods of each payment case to obtain the total treatment period when there are multiple insurance payout cases with the same disease classification code and different reference dates.

[0017] In an automated insurance contract review process utilizing a significant disease information judgment and risk classification model according to the present invention, the mandatory disclosure disease information selection unit is characterized by generating the mandatory disclosure disease information DB by determining whether each coverage is mandatory based on the disease classification code by combining the stored values ​​of an underwriting standard DB in which disease classification codes, insurance payout amounts, treatment details, treatment periods, elapsed time after the end of treatment, and disease review guidelines for each coverage coverage of the insurance are mapped, the customer's application information, and variable values ​​included in the health information DB.

[0018] The insurance contract review process automation device utilizing a significant disease information judgment and risk classification model according to the present invention is characterized by further including a modular disease scenario query unit that retrieves essential disclosure disease information stored in the essential disclosure disease information DB and performs a query on the essential disclosure disease information by referring to a modular disease scenario DB in which a query-answer set corresponding to the retrieved essential disclosure disease information is stored modularly.

[0019] The insurance contract review process automation device utilizing a significant disease information judgment and risk classification model according to the present invention is characterized by further including a coverage-specific guidance message providing unit, wherein the modular disease scenario query unit performs a query regarding the mandatory disclosure disease information by referring to the modular disease scenario DB, and when the underwriting result is determined according to the customer's response to the query, the unit refers to a coverage-specific guidance message DB in which guidance messages regarding the underwriting result are stored by coverage and provides guidance messages regarding the underwriting result by classifying them by coverage. Effects of the invention

[0020] According to the present invention, there is an effect of providing an automated insurance contract review process utilizing a significant disease information judgment and risk classification model, which enables insurance policyholders and insurance solicitors to check the conditions for and eligibility of an insurance contract in real time, drastically reduce the number of mandatory disclosure disease items, and allow insurers to operate an efficient insurance contract review organization by simplifying the insurance application process. Brief explanation of the drawing

[0021] FIG. 1 is a drawing showing an insurance contract review process automation device utilizing a significant disease information judgment and risk classification model according to an embodiment of the present invention, and FIG. 2 is a diagram exemplarily illustrating the operational configuration of a significant disease information extraction unit in one embodiment of the present invention, and FIGS. 3 to 12 are drawings for specifically and exemplarily explaining the process of a significant disease information extraction unit generating a health information DB, and FIG. 13 is a diagram exemplarily illustrating the operational configuration of a mandatory notification disease information selection unit in one embodiment of the present invention, and FIGS. 14 to 18 are drawings for specifically and exemplarily explaining the process by which a mandatory notification disease information selection unit generates a mandatory notification disease information DB, and FIG. 19 is a diagram exemplarily illustrating the operational configuration of a recommendation unit for coverage eligible for subscription tailored to individual medical history in one embodiment of the present invention, and FIGS. 20 and 21 are diagrams intended to specifically and exemplarily explain the process of a recommendation section for coverage eligible for personalized medical history recommending coverage eligible for personalized medical history. FIG. 22 is a diagram exemplarily illustrating the operational configuration of a modular disease scenario query unit in one embodiment of the present invention, and FIGS. 23 to 25 are drawings for specifically and exemplarily explaining the process of a modular disease scenario query unit querying a disease using a modular disease scenario DB, and FIG. 26 is a diagram exemplarily illustrating the operational configuration of a guidance message providing unit by collateral in one embodiment of the present invention, and FIGS. 27 and 28 are drawings intended to specifically and exemplarily explain the process by which a collateral-specific guidance message providing unit provides collateral-specific guidance messages using a collateral-specific guidance message DB, and FIG. 29 is a diagram illustrating an exemplary insurance contract application process using an insurance contract review process automation device utilizing a significant disease information judgment and risk classification model according to one embodiment of the present invention. Specific details for implementing the invention

[0022] Since embodiments according to the concept of the present invention may be subject to various modifications and may take various forms, embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the concept of the present invention to specific disclosed forms, and includes all modifications, equivalents, or substitutions that fall within the spirit and scope of the present invention.

[0023] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an ideal or overly formal sense unless explicitly defined herein.

[0024] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the attached drawings.

[0025] FIG. 1 is a drawing showing an insurance contract review process automation device (40) utilizing a significant disease information judgment and risk classification model according to one embodiment of the present invention, and FIG. 29 is a drawing showing an overall insurance contract application process using an insurance contract review process automation device (40) utilizing a significant disease information judgment and risk classification model according to one embodiment of the present invention in an exemplary and specific manner.

[0026] Referring to FIG. 1 and FIG. 29, an insurance contract review process automation device (40) utilizing a significant disease information judgment and risk classification model according to one embodiment of the present invention may be configured to include a significant disease information extraction unit (100), a mandatory disclosure disease information selection unit (200), a personal disease history customized eligible coverage recommendation unit (300), a modular disease scenario query unit (400), and a coverage-specific guidance message provision unit (500).

[0027] The significant disease information extraction unit (100) performs the function of extracting significant disease information of a customer from the customer's insurance payment information accumulated in the credit information concentration agency (30) and storing the extracted significant disease information in the health information DB (110).

[0028] For example, the significant disease information extraction unit (100) can determine the validity of each treatment content corresponding to each payment information by using the disease classification code (KCD), treatment content (surgery, hospitalization, outpatient), treatment start date, treatment end date, date of occurrence of insurance accident, date of payment of insurance, number of treatments, number of treatment days, and total treatment period included in the insurance payment information message received from the credit information concentration agency (30), and can determine a reference date to be used for calculating the elapsed period by applying a pre-set reference date determination rule to the treatment content determined to be valid.

[0029] For example, if there are multiple reference dates for the same disease classification code, the significant disease information extraction unit (100) may use the reference date closest to the current time point to calculate the elapsed period.

[0030] For example, the significant disease information extraction unit (100) may assign a pre-set priority to treatment content corresponding to each payment information included in the insurance payment information provided by the credit information concentration agency (30), and if multiple treatment contents are valid, select one treatment content according to the priority to create a health information DB (110).

[0031] For example, if there are multiple treatment contents for the same disease classification code, the significant disease information extraction unit (100) can create a health information DB (110) by selecting one treatment content according to priority.

[0032] For example, if there are multiple insurance payout cases with the same disease classification code and reference date, the significant disease information extraction unit (100) can generate a health information DB (110) by selecting the highest amount for the payout, selecting the highest number of times for each treatment content for the number of treatments, selecting the maximum number of days for the treatment, and selecting the maximum period among each payout case for the total treatment period.

[0033] For example, if there are multiple insurance payout cases with the same disease classification code and different reference dates, the significant disease information extraction unit (100) can generate a health information DB (110) by summing the respective payout amounts, summing the number of treatments and treatment days by treatment content, and summing the treatment periods of each payout case for the total treatment period.

[0034] With further reference to FIGS. 2 to 12, the configuration in which the significant disease information extraction unit (100) generates the health information DB (110) is described in detail and exemplarily as follows.

[0035] FIG. 2 is a diagram illustrating the operational configuration of a significant disease information extraction unit (100) in one embodiment of the present invention, and FIG. 3 to FIG. 12 are diagrams for specifically and exemplarily explaining the process of the significant disease information extraction unit (100) generating a health information DB (110).

[0036] Referring further to FIGS. 2 to 12, the credit information concentration agency (30) may be the Korea Credit Information Service, and the Korea Credit Information Service’s integrated insurance credit information inquiry system provides insurance credit information, which is information necessary for determining the transaction details of a credit information subject for the purpose of preventing insurance fraud in relation to insurance commercial transactions such as the conclusion of insurance contracts and the claim and payment of insurance benefits. Such insurance credit information may include information regarding the conclusion of insurance contracts and information regarding the claim and payment of insurance benefits. Information regarding the conclusion of insurance contracts may include information regarding the status of insurance contracts, information regarding the insured or the claimant of insurance benefits, and information regarding the solicitation business trustee. Information regarding the status of insurance contracts may include the insurance contract date, insurance period, name of the insurance product, name of the benefit, sum insured, premium, name of the insurance company and insurance contract status, information on the insured object, application information, etc. Information regarding the insured or the claimant of the insurance contract may include information regarding the name, personal identification number, occupation, and relationship with the policyholder, etc. Information regarding the solicitation business trustee may include the name, designation, and registration number, etc. of the solicitation business trustee who solicited the relevant insurance contract.

[0037] The significant disease information extraction unit (100) extracts necessary items from the information included in the insurance payment information provided by the credit information concentration agency (30) to extract the customer's significant disease information, and stores the extracted customer's significant disease information in the health information DB (110).

[0038] The reason the significant disease information extraction unit (100) generates the health information DB (110) is as follows.

[0039] If there is multiple insurance payout information with the same diagnosis on the same accident date, it is necessary to organize duplicate data by distinguishing between ① and ②.

[0040] ① Whether it constitutes duplicate billing to multiple companies for a single treatment of the same disease

[0041] ② Whether it constitutes a recurrence or additional treatment for the same disease

[0042] In other words, if data is called out redundantly despite having the same medical history, all of it must be disclosed, or it may be classified as an insurance claim payment case (medical history) that the contract underwriter must verify, requiring an additional process of requesting the customer to re-disclose it.

[0043] Therefore, a technology that can refine the information included in the insurance payout information provided by the credit information concentration agency (30) and link it with the automation of the examination process is emerging as a key element of the insurance company's new contract subscription system. To this end, it is necessary to develop a technology that can adjust the information provided by the credit information concentration agency regarding treatment details (type, priority), total treatment period (number of treatments, number of treatment days), and treatment dates (date of original accident, start date of treatment, end date of treatment).

[0044] Figure 3 illustrates raw data of insurance payout information provided by a credit information concentration agency (30), and Figure 4 illustrates a health information DB (110) generated by a significant disease information extraction unit (100) processing the raw data of insurance payout information illustrated in Figure 3. The significant disease information extraction unit (100) can be configured to adjust raw data of treatment details (type, priority), total treatment period (number of treatments, number of treatment days), and treatment date and time (date and time of original accident, treatment start date, treatment end date) to extract significant disease information of various insurance payout cases, such as duplicate claims for the same disease and same treatment, and additional claims for the same disease and additional treatment, and to store the extracted significant disease information in the health information DB (110).

[0045] A specific and exemplary method for generating a health information DB (110) by extracting significant disease information is described as follows with reference to FIGS. 5 to 12.

[0046] (1) Basic assumption

[0047] If treatment is defined by classifying it into Treatment A, Treatment B, and Treatment C according to the treatment content, the value of Treatment A is T if the number of Treatment A is greater than 0, and the value of Treatment A is F if the number of Treatment A is 0; the value of Treatment B is T if the number of days of Treatment B is greater than 0 or the number of Treatment B is greater than 0; the value of Treatment B is F if the number of days of Treatment B is 0 and the number of Treatment B is 0; the value of Treatment C is T if the number of Treatment C is 0, and the value of Treatment C is F if the number of Treatment C is 0; and assuming the total treatment period is SUM [number of days of Treatment B, number of Treatment A], that is, the sum of the number of days of Treatment B and the number of Treatment A, the result exemplified in Fig. 5 can be obtained. Treatment C is excluded as it is irrelevant to the calculation of the treatment period due to the characteristics of the data.

[0048] (2-1) Basics: Definition of Reference Date Determination Rules

[0049] Using the values ​​for the number of treatments A, B, and C and the number of days of treatment B corresponding to each payment information of the insurance payment information, the reference date to be used for calculating the elapsed period is processed as exemplified in Fig. 6 below.

[0050] (2-2) Addition: Additional definition of reference date determination rule

[0051] If there are multiple identical disease classification codes, the following criteria are additionally applied to delete duplicate data, and then the health information DB (110) is finally created and stored.

[0052] That is, when determining the reference date, if multiple reference dates exist for the same disease classification code, the rule Max[Reference Date 1, Reference Date 2, … , Reference Date N] is applied. For example, in the case of Fig. 7, end date 9 is selected according to condition ④ of Fig. 5. In this case, the reference date is the date closest to the current time.

[0053] (3-1) Basics: Definition of Treatment Processing Methods

[0054] For the treatment details (treatment A, treatment B, treatment C) present in the insurance payment information, a priority standard for the treatment details as exemplified in FIG. 8 is established to create a health information DB (110).

[0055] (3-2) Addition: Additional definition of treatment processing method

[0056] Regarding treatment content, if multiple treatment contents exist under the same disease classification code, the criteria "(3-1) Basic: Definition of treatment content processing method" are reapplied as exemplified in Fig. 9.

[0057] (4) Number of Treatment A, Number of Treatment B Days, Number of Treatment B, Number of Treatment C, Total Treatment Period, Determination of Payment Amount

[0058] 1) Cases where the disease classification code and reference date are identical and there are multiple payment cases

[0059] The number of treatments A is MAX [Treatment A count1, Treatment A count2, … , Treatment A countN]. In other words, the largest value for the number of treatments A is selected. Likewise, the number of days for treatment B is MAX [Treatment B days1, Treatment B days2, … , Treatment B daysN], the number of treatments B is MAX [Treatment B count1, Treatment B count2, … , Treatment B countN], the number of treatments C is MAX [Treatment C count1, Treatment C count2, … , Treatment C countN], the total treatment period is MAX [Total treatment period1, Total treatment period2, … , Total treatment periodN], and the payment amount is MAX [Payment amount1, Payment amount2, … , Payment amountN].

[0060] In cases where the disease classification code and reference date are the same and there are multiple payment cases, the method for determining the number of treatment A, the number of treatment B days, the number of treatment B, the number of treatment C, the total treatment period, and the payment amount is exemplified in FIG. 10.

[0061] 2) Cases where the disease classification code is identical, the reference dates differ, and there are multiple payment cases

[0062] The number of treatments A is SUM [Treatment A count1, Treatment A count2, … , Treatment A countN]. That is, the number of treatments A is the sum of all treatments A counts. Likewise, the number of treatment B days is SUM [Treatment B days1, Treatment B days2, … , Treatment B daysN], the number of treatments B is SUM [Treatment B count1, Treatment B count2, … , Treatment B countN], the number of treatments C is SUM [Treatment C count1, Treatment C count2, … , Treatment C countN], the total treatment period is SUM [Total treatment period1, Total treatment period2, … , Total treatment periodN], and the payment amount is SUM [Payment amount1, Payment amount2, … , Payment amountN].

[0063] Figure 11 illustrates the method for determining the number of treatments A, the number of days of treatment B, the number of treatments B, the number of treatments C, the total treatment period, and the payment amount in cases where the disease classification code is the same, the reference date is different, and there are multiple payment cases.

[0064] Through the above exemplary description, a health information DB (110) generated from insurance payment information data by a significant disease information extraction unit (100) is exemplified in FIG. 13.

[0065] The mandatory notification disease information selection unit (200) performs the function of selecting mandatory notification disease information by applying the insurance company's disease-specific underwriting criteria information to significant disease information stored in the health information DB (110), and storing the selected mandatory notification disease information in the mandatory notification disease information DB (210).

[0066] For example, the mandatory disclosure disease information selection unit (200) can generate a mandatory disclosure disease information DB (210) by determining whether to make a mandatory disclosure for each coverage based on the disease classification code by combining variable values ​​included in the underwriting standard DB (220), which maps the disease classification code, insurance payout amount, treatment details, treatment period, elapsed time after treatment end, and disease screening guidelines for each coverage of the insurance, and the customer's subscription information and health information DB (110).

[0067] With further reference to FIGS. 13 to 18, the configuration in which the mandatory notification disease information selection unit (200) generates the mandatory notification disease information DB (210) is described in detail and exemplarily as follows.

[0068] FIG. 13 is a diagram illustrating the operational configuration of a mandatory notification disease information selection unit (200) in one embodiment of the present invention, and FIG. 14 to 18 are diagrams for specifically and exemplarily explaining the process of the mandatory notification disease information selection unit (200) generating a mandatory notification disease information DB (210).

[0069] Referring further to FIGS. 13 and 14, the mandatory notification disease information selection unit (200) extracts necessary items from the information included in the health information DB (110), selects mandatory notification disease information using the information stored in the acceptance standard DB (220), the Korean Standard Classification of Diseases and Causes of Death (KCD), standard information related to disease names, and the customer's subscription information, and stores the selected mandatory notification disease information in the mandatory notification disease information DB (210).

[0070] Figure 15 illustrates an example of the components of an acceptance standard DB (220) that includes items such as payment amount, treatment method, treatment period, and elapsed time after treatment type.

[0071] For example, in order to select mandatory notification diseases by calling columns and data within the health information DB (110), a logic is required to correspond the values ​​mapped to each column to the result, and the acceptance standard DB (220) can be utilized for this logic. For example, the acceptance standard DB (220) can be created by converting a disease screening guideline in the form of a document, expressed in a general user-centric manner and including explanatory materials on the characteristics of each disease, into a DB format.

[0072] Referring further to FIGS. 16 to 18, the method by which the mandatory notification disease information selection unit (200) selects mandatory notification disease information is explained in a specific and exemplary manner.

[0073] Figure 16 shows examples of mandatory disclosure items required based on insurance payment information data, Figure 17 shows examples of mandatory disclosure items required based on a health information DB (110) configured based on significant disease information extracted from the customer's insurance payment information, and Figure 18 shows examples of mandatory disclosure items required for each coverage based on a mandatory disclosure disease information DB (210).

[0074] Referring to FIGS. 16 to 18, if there is no mandatory disclosure disease information DB (210), there is an inconvenience in having to disclose all primary and secondary diagnosis disease classification codes found in the insurance payment information details. As a specific example, even minor matters such as one day of cold treatment or one day of skin boil removal must be disclosed.

[0075] However, it is difficult for actual customers to remember all diseases, and for diseases that are accepted even if disclosed, it is important to simplify the application process by minimizing unnecessary disclosure of medical history within the duty to disclose prior to the contract.

[0076] Meanwhile, even if the health information DB (110) according to one embodiment of the present invention is used, detailed acceptance criteria for each disease are not provided, so it is still necessary to disclose medical history for all insurance payout capabilities.

[0077] However, as illustrated in FIG. 18, by using the mandatory disclosure disease information DB (210) based on the combination of subscription information and health information DB data, it can be seen that the scope of disclosure for all coverages can be drastically reduced, and the results can be verified for each coverage that the policyholder wishes to subscribe to.

[0078] The personal medical history customized coverage recommendation unit (300) performs the function of recommending personal medical history customized coverage to the customer based on the customer's subscription information, significant disease information stored in the health information DB (110), and mandatory disclosure disease information stored in the mandatory disclosure disease information DB (210).

[0079] With additional reference to FIGS. 19 to 21, the operation of the personal disease history customized coverage recommendation unit (300) recommending personal disease history customized coverage to a customer is described in detail and exemplarily as follows.

[0080] FIG. 19 is a diagram illustrating the operational configuration of a personal disease history customized coverage recommendation unit (300) in one embodiment of the present invention, and FIG. 20 and FIG. 21 are diagrams for specifically and exemplarily explaining the process of the personal disease history customized coverage recommendation unit (300) recommending personal disease history customized coverage.

[0081] Referring to the example in Fig. 20, there is a problem in that, as with conventional technology, if there is no personalized recommendation function for available coverage, one must design all coverages (special riders) and amounts of the desired product and go through a complex contract review process to check the underwriting results, and if the final result is deemed unacceptable, one must request a redesign and undergo a contract review process excluding the coverages in question.

[0082] However, referring to the example in FIG. 21, if there is a personalized coverage recommendation function as in one embodiment of the present invention, coverage available for subscription can be identified first at the time of starting product design, thereby simplifying complex contract subscription processes such as redesign and re-examination, which were problems of the existing process.

[0083] As illustrated in FIG. 21, one embodiment of the present invention shows results for all coverages in operation, allows for the selection of only coverages that are eligible for acceptance to proceed with subscription design and application, and can infinitely expand the scope to derive results for all coverages in operation by the insurance company. In other words, there is no limit to the number of coverages. In FIG. 21, "eligible for acceptance" means that acceptance is possible immediately without conditions, "requires examination" means that acceptance is possible after examination as it is subject to mandatory disclosure, and "uneligible for acceptance" means rejection.

[0084] The modular disease scenario query unit (400) has the function of calling the mandatory disclosure disease information stored in the mandatory disclosure disease information DB (210), and performing a query on the mandatory disclosure disease information by referring to the modular disease scenario DB (410), in which a set of questions and answers corresponding to the called mandatory disclosure disease information is stored in a modular manner. After the customer performs the query and answer on the mandatory disclosure disease information, the insurance agent, etc., can check the results of the query and answer on the mandatory disclosure disease information and the underwriting results by coverage by referring to the disease screening guidelines by coverage stored in the underwriting criteria DB (220).

[0085] With further reference to FIGS. 22 to 25, the operation of the modular disease scenario query unit (400) performing a query on the customer regarding mandatory notification disease information is described in detail and exemplarily as follows.

[0086] FIG. 22 is a diagram illustrating the operational configuration of a modular disease scenario query unit (400) in an embodiment of the present invention, and FIG. 23 to 25 are diagrams for specifically and exemplarily explaining the process of the modular disease scenario query unit (400) querying a disease using a modular disease scenario DB (410).

[0087] Referring to the example in Fig. 24, the conventional technology utilizes a responsive disease scenario DB in which the next question is identified based on the answer to each disease-specific question. According to this method, questions are configured based on each disease classification code (KCD), making modifications and changes difficult. Furthermore, since both disease A and disease B operate with identical questions and answers, if the questions and answers need to be modified, the entire responsive scenario for disease A and disease B must be modified. Additionally, the disease scenario must be changed immediately upon changes to the disease screening guidelines that form the basis of the disease scenario DB, but there is a problem in that the scope of change for the responsive disease scenario is wide, making immediate changes difficult.

[0088] However, referring to the example in FIG. 25, one embodiment of the present invention performs a query using a modular disease scenario DB (410) composed of a question-answer set. For example, questions 1 and 2 are configured in common, and up to 5 additional questions are possible. Since only the code of the question-answer set needs to be mapped by KCD, it is easy to manage and has a modular structure that allows for various changes according to desired combinations, which has the effect of being differentiated from the prior art.

[0089] The coverage-specific guidance message provision unit (500) performs the function of providing guidance messages regarding the acquisition result by distinguishing them by coverage when the modular disease scenario query unit (400) performs a query for mandatory notification disease information by referring to the modular disease scenario DB (410), and when the acquisition result is determined according to the customer's answer to this query, by referring to the coverage-specific guidance message DB (510) in which guidance messages regarding the acquisition result are stored by coverage.

[0090] With further reference to FIGS. 26 to 28, the operation of the guidance message providing unit (500) for each collateral providing guidance message is described in detail and exemplarily as follows.

[0091] FIG. 26 is a diagram illustrating the operational configuration of a collateral-specific guidance message providing unit (500) in an embodiment of the present invention, and FIG. 27 and FIG. 28 are diagrams for specifically and exemplarily explaining the process of the collateral-specific guidance message providing unit (500) providing a collateral-specific guidance message using a collateral-specific guidance message DB (510).

[0092] Referring to the example in FIG. 27, a form of guidance message according to the prior art is disclosed. Although detailed guidance messages are required for each reason for the output of the contract review results for each collateral, according to the prior art, there is no guidance message DB (510) for each collateral, so there is no guidance other than the contract review results. In addition, regarding the main contract and CI collateral in the case exemplified in FIG. 27, detailed guidance (attachment of documents, guidance on diagnosis, need for underwriter review) is required to proceed with the review, but since no guidance message is output, there is a problem that for such contract review results, one must make an individual inquiry to the contract reviewer to check the details in order to proceed with the subsequent contract review process. Furthermore, even in the case of non-acceptance, guidance on the minimum elapsed period for re-review is required, but the prior art fails to address this.

[0093] However, referring to the example in FIG. 28, one embodiment of the present invention can provide detailed information such as the type of document, the validity period of the document, and the minimum waiting period for re-examination for each type of collateral. Through this, the user of the subscription process (insurance agent) can provide detailed guidance to the customer, such as the need for additional information including the final contract review results and document submission, thereby improving the sales support function.

[0095] As explained in detail above, according to the present invention, there is an effect of providing an insurance contract review process automation device (40) utilizing a significant disease information judgment and risk classification model that allows the insurance policyholder and the insurance agent to check the conditions for subscription and eligibility for subscription of the insurance contract in real time, drastically reduce the number of items of mandatory disclosure diseases, and enable the insurer to operate an efficient insurance contract review organization through the simplification of the insurance application process.

[0096] More specifically, according to one embodiment of the present invention, when using a significant disease information extraction technique and a mandatory disclosure disease screening technique, the customer's insurance payment information is automatically retrieved through a credit information agency, so that necessary information can be automatically retrieved even if the customer does not remember and input all of their medical history.

[0097] Furthermore, during the disease disclosure stage of the life insurance application process, a customer's medical history can be retrieved from credit information agencies that collect data on payout capacity to assist with the disclosure. However, if the entire customer's medical history is retrieved from these agencies and requires a detailed disclosure of treatment history, inefficiency may arise. In cases of minor illnesses that do not pose a mortality risk or affect contract acceptance, underwriting may be possible even without the mandatory disclosure of detailed treatment history. However, if the entire medical history is retrieved, disclosure becomes unavoidable, which can become a cumbersome procedure for both the customer and the insurance company.

[0098] One embodiment of the present invention improves upon this by generating a health information DB (110) through a significant disease information extraction technique and applying a mandatory disclosure disease selection technique so that the review results can be checked even if only relevant insurance payment information is disclosed. The customer can automatically check the final review results by entering their treatment history in detail according to a predetermined disease review scenario, only for the medical history retrieved through the final generated mandatory disclosure disease information DB (210).

[0099] Furthermore, the significant disease information extraction technique and the mandatory disclosure disease screening technique according to one embodiment of the present invention are not merely utilized for detailed disease disclosure, but can also expand their scope of application to include product proposals. Based on the underwriting results for each coverage type retrieved from the customer's insurance claim payment history, it is possible to propose eligible coverages; this enables product recommendations and coverage (special rider) recommendations tailored to the individual insured's medical history. Previously, the eligibility for products and individual coverages could only be confirmed after completing both insurance design and screening; if a final result indicated ineligibility, redesign and re-screening were required. However, one embodiment of the present invention significantly reduces these inconveniences and enables the simplification of the entire subscription process.

[0100] In addition, through the significant disease information extraction technique and the mandatory disclosure disease screening technique according to one embodiment of the present invention, the acceptance of medical history requiring final disclosure is determined after detailed treatment history disclosure is performed. The detailed treatment history disclosure method used at this time consists of questions and answers pre-set by the insurance company.

[0101] Furthermore, prior to one embodiment of the present invention, the question-and-answer method was structured in the form of a responsive disease scenario, where the next question was identified based on the answer provided for each disease. Under this structure, there is a disadvantage in that it is not easy to modify or change the scenarios. For example, if the same query called Q1 is operated for both disease code A and disease code B, and a change to the content of Q1 is required, modification work must be performed separately in the scenario databases for both disease code A and disease code B. Recently, as part of marketing efforts to increase sales volume, insurance companies have been temporarily relaxing or adjusting disease screening guidelines; at this time, changes to responsive disease scenarios for disclosing detailed treatment history are also frequently required. Since there are approximately 20,000 total Korean Classification of Diseases (KCD) codes in Korea, the scope of modification can be substantial if multiple changes are made.

[0102] One embodiment of the present invention adopts a method of creating a modular disease scenario DB (410) to compensate for the above disadvantages and retrieving necessary queries and answers in accordance with KCD. By analyzing the previously operated responsive disease scenarios, items commonly used in various diseases are classified as common items, and scenarios are configured in the form of additional items so that items can be operated differently according to the characteristics of each disease.

[0103] In other words, query items and answers were created in a modular disease scenario database, and they were mapped so that they could be retrieved from the database for each disease code. Consequently, when changes are required, modifying only the content of the query item or answer automatically reflects the changes across all disease codes mapped to them, significantly reducing the scope of modifications and enhancing ease of management.

[0104] In addition, in one embodiment of the present invention, the results of the contract review can be verified by inputting detailed information about the disease through the query-answer generated via the modular disease scenario DB. The results of the contract review are classified into standard, substandard (conditional acceptance such as surcharge or exclusion), unacceptable, underwriter review required, and additional information required. In the case of standard and substandard results, customer notification regarding the contract review results is not required, but customer notification is required for all other results. If an unacceptable result is confirmed, the minimum period for re-examination of acceptance must be provided; if an underwriter review is required or additional information is needed, the customer must be notified of the medical documents or necessary health checkup items that must be submitted.

[0105] Prior to one embodiment of the present invention, only the final result of the contract review was provided without any notification message regarding the contract review results. In other words, although the disease review guidelines were structured differently for each coverage type, it was difficult to immediately provide information that actual recruiters and customers were curious about, such as the medical documents and diagnostic items required for re-review for each coverage type, and the minimum waiting period for re-review. This caused inefficiency, as specific information could only be obtained by making inquiries to the contract reviewer.

[0106] One embodiment of the present invention is configured with guidance messages by type of collateral to compensate for these disadvantages, and this is the biggest difference from existing guidance messages. Through the guidance message DB (510) by type of collateral, detailed information such as the types of documents, validity periods of documents, and minimum waiting periods required for re-examination can all be provided. Through this, a more detailed guidance function regarding the contract review results is implemented, which also helps the sales organization with smooth sales activities.

[0107] Through the aforementioned features, policyholders and insurance agents can verify the terms and eligibility of insurance contracts in real time. Ultimately, by expanding the proportion of approved contracts through an automated underwriting process based on techniques developed by our company, insurers can reduce costs associated with the application process and operate a highly efficient insurance contract underwriting organization. Explanation of the symbols

[0108] 10: Insurance agent terminal 20: Customer terminal 30: Credit Information Bureau 40: Insurance contract underwriting process automation device 100: Significant disease information extraction unit 110: Health Information DB 200: Mandatory Notification Disease Information Screening Section 210: Mandatory Disclosure Disease Information DB 220: Acceptance Criteria DB 300: Recommendation for coverage tailored to individual medical history 400: Modular Disease Scenario Inquiry Section 410: Modular Disease Scenario DB 500: Guidance Message Provision Department by Collateral 510: Guidance Message DB by Collateral

Claims

Claim 1 An automated insurance contract review process utilizing a model for determining significant disease information and classifying risk, comprising: a significant disease information extraction unit that extracts significant disease information of a customer from the customer's insurance claim payment information accumulated at a credit information concentration agency and stores the extracted significant disease information in a health information DB; and a mandatory disclosure disease information selection unit that applies the insurer's underwriting criteria information to the significant disease information stored in the health information DB to select mandatory disclosure disease information and stores the selected mandatory disclosure disease information in a mandatory disclosure disease information DB. The system includes a personal medical history customized coverage recommendation unit that recommends coverages eligible for enrollment tailored to the personal medical history to the customer based on the customer's subscription information, significant disease information stored in the health information DB, and mandatory disclosure disease information stored in the mandatory disclosure disease information DB. The significant disease information extraction unit determines the validity of each treatment content corresponding to each payment information using the disease classification code (KCD), treatment details (surgery, hospitalization, outpatient), treatment start date, treatment end date, date of insured event, date of insurance payment, number of treatments, number of treatment days, and total treatment period corresponding to each payment information included in the insurance payment information message received from the credit information concentration agency; determines a reference date to be used for calculating the elapsed period by applying a pre-set reference date determination rule to the treatment content determined to be valid; if multiple reference dates exist for the same disease classification code, uses the reference date closest to the current time point for calculating the elapsed period; assigns a pre-set priority to each treatment content corresponding to each payment information included in the insurance payment information message; and if multiple treatment contents are valid, one according to the priority An insurance contract review process automation device that generates the above health information DB by selecting treatment content, wherein if multiple treatment contents exist for the same disease classification code, one treatment content is selected according to the above priority to generate the above health information DB. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete Claim 6 An insurance contract review process automation device according to claim 1, wherein the above-mentioned significant disease information extraction unit generates the above-mentioned health information DB by selecting the highest amount for the payment amount, selecting the highest number of times for each treatment content for the number of treatments, selecting the maximum number of days for the treatment, and selecting the maximum period among each payment case for the total treatment period when there are multiple insurance payment cases with the same disease classification code and reference date. Claim 7 An insurance contract review process automation device according to claim 1, wherein the above-mentioned significant disease information extraction unit generates the above-mentioned health information DB by summing the respective payment amounts, summing the number of treatments and treatment days by treatment content respectively, and summing the treatment periods of each payment case for the total treatment period when there are multiple insurance payment cases with the same disease classification code and different reference dates. Claim 8 An insurance contract review process automation device according to claim 1, wherein the mandatory disclosure disease information selection unit generates the mandatory disclosure disease information DB by determining whether to make a mandatory disclosure for each coverage based on the disease classification code by combining the stored values ​​of an underwriting standard DB in which disease classification codes, insurance payout amounts, treatment details, treatment periods, elapsed time after the end of treatment, and disease review guidelines for each coverage coverage of the insurance are mapped, the customer's application information, and variable values ​​included in the health information DB. Claim 9 An insurance contract review process automation device according to claim 1, further comprising a modular disease scenario query unit that retrieves mandatory disclosure disease information stored in the mandatory disclosure disease information DB and performs a query on the mandatory disclosure disease information by referring to a modular disease scenario DB in which a query-answer set corresponding to the retrieved mandatory disclosure disease information is stored modularly. Claim 10 An insurance contract review process automation device according to claim 9, further comprising a coverage-specific guidance message providing unit that, when the modular disease scenario query unit performs a query regarding the mandatory disclosure disease information by referring to the modular disease scenario DB and the acceptance result is determined according to the customer's response to the query, provides guidance messages regarding the acceptance result by classifying them by coverage by referring to a coverage-specific guidance message DB in which guidance messages regarding the acceptance result are stored by coverage.

Images