JP7561114B2 - Information analysis support method and information analysis support system - Google Patents

Description

The present invention relates to technology that supports the analysis of information related to health conditions.

Life insurance companies determine the risk of insurance payment based on the policyholder's health condition at the time of contracting the policy and perform underwriting assessments. The policyholder's health condition at the time of contracting the policy is identified based on notification information submitted when applying for insurance. One example of technology for supporting the input of notification information is described in JP 2015-22374 (Patent Document 1). Patent Document 1 describes, "The computer is caused to execute a process of acquiring medical information corresponding to personal identification information that identifies a user, a process of selecting question text information and answer item information corresponding to the medical information from question text information and answer item information corresponding to the notification items in the insurance product based on the items of the medical information, and a process of outputting input screen data for the notification items including the selected question text information and answer item information."

JP 2015-22374 A

The information submitted by applicants for membership includes many items indicating their health condition, and may contain ambiguous expressions, typos, or incorrect information. As a result, data preparation is done manually in advance, and it takes a lot of time and money to reach the point of risk assessment. In order to perform risk assessment more accurately and easily, it is necessary to make it possible to carry out this process automatically.

The above-mentioned Patent Document 1 describes a technology that supports the entry of information by applicants based on medical records, but does not disclose a technology for converting the submitted information into a format suitable for risk assessment.

To solve the above problems, the present invention aims to provide a method for extracting and organizing information necessary for information processing such as risk assessment based on the health status information entered by the applicant, and for enabling information processing using a specified model, etc.

In order to solve at least one of the above problems, a representative example of the invention disclosed in this application is an information analysis support method executed by a computer system, the computer system having a processor and a storage device connected to the processor, the storage device holding a risk model for calculating a risk value related to health, dictionary information for associating health information with at least one of a plurality of aggregation categories for aggregating the health information, and explanatory variable definition information for associating one or more combinations of the aggregation categories with values of explanatory variables of the risk model, the information analysis support method including a first step in which the processor extracts the health information from input information, a second step in which the processor identifies one or more aggregation categories corresponding to the extracted health information based on the dictionary information, a third step in which the processor identifies the values of the explanatory variables corresponding to the identified combinations of the one or more aggregation categories based on the explanatory variable definition information, and a fourth step in which the processor calculates the risk value corresponding to the identified values of the explanatory variables based on the risk model.

According to one aspect of the present invention, the time and cost required for processing is reduced by automating data preparation up to information processing such as risk assessment. Problems, configurations, and effects other than those described above will be made clear by the explanation of the following embodiment.

1 is a block diagram showing an example of a configuration of an information analysis support system according to a first embodiment of the present invention. 1 is an explanatory diagram illustrating an example of basic information managed by a basic information management unit according to the first embodiment of the present invention; 1 is an explanatory diagram showing an example of notification information managed by a health information management unit according to the first embodiment of the present invention; FIG. 1 is an explanatory diagram showing an example of health check information managed by a health information management unit according to the first embodiment of the present invention; 1 is an explanatory diagram showing an example of disease name definition information managed by a disease name definition information management unit according to the first embodiment of the present invention; 1 is an explanatory diagram showing an example of drug definition information managed by a drug definition information management unit according to the first embodiment of the present invention; 1 is an explanatory diagram showing an example of surgery definition information managed by a surgery definition information management unit in Example 1 of the present invention. FIG. 1 is an explanatory diagram showing an example of hospitalization definition information managed by a hospitalization definition information management unit according to a first embodiment of the present invention; 1 is an explanatory diagram illustrating an example of test result definition information managed by a test result definition information management unit according to the first embodiment of the present invention; FIG. 1 is an explanatory diagram illustrating an example of explanatory variable definition information managed by an explanatory variable definition information management unit according to the first embodiment of the present invention; 11 is an explanatory diagram illustrating an example of an intermediate table of the explanatory variable generation process managed by an explanatory variable generation result management unit according to the first embodiment of the present invention. FIG. 1 is an explanatory diagram illustrating an example of risk calculation target information managed by a risk calculation target information management unit according to the first embodiment of the present invention; 11 is an explanatory diagram illustrating an example of explanatory variable generation result information managed by an explanatory variable generation result management unit according to the first embodiment of the present invention. FIG. An explanatory diagram showing an example of risk model information managed by the risk model information management unit of Example 1 of the present invention. 1 is an explanatory diagram illustrating an example of classification definition information managed by a classification definition information management unit according to the first embodiment of the present invention; 11 is a flowchart illustrating an example of a process executed by a disease name information extraction unit according to the first embodiment of the present invention. 11 is a flowchart showing an example of a process executed by a drug information extraction unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of a process executed by a test result information extraction unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of processing executed by a surgery information extraction unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of a process executed by a hospitalization information extraction unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of processing executed by an explanatory variable information generating unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of a process executed by a risk calculation unit according to the first embodiment of the present invention. 11 is a flowchart illustrating an example of a process executed by a risk determination unit according to the first embodiment of the present invention. 11 is an explanatory diagram illustrating an example of a user interface that displays notification information in the process of generating explanatory variables by the information analysis support system according to the first embodiment of the present invention; FIG. 11 is an explanatory diagram illustrating an example of a user interface that displays a result of an aggregation process in the process of generating explanatory variables by the information analysis support system according to the first embodiment of the present invention. FIG. 1 is an explanatory diagram illustrating an example of a user interface that displays results of an explanatory variable generation process in the process of generating explanatory variables by the information analysis support system according to the first embodiment of the present invention; FIG. 1 is an explanatory diagram illustrating an example of a user interface that displays a result of a risk analysis performed by the information analysis support system according to the first embodiment of the present invention. FIG. FIG. 11 is a block diagram showing an example of a configuration of an information analysis support system according to a second embodiment of the present invention. FIG. 11 is an explanatory diagram illustrating an example of contract information managed by a contract information management unit according to the second embodiment of the present invention. FIG. 11 is an explanatory diagram illustrating an example of billing information managed by a billing information management unit according to the second embodiment of the present invention. An explanatory diagram showing an example of payment division definition information managed by the payment division definition information management unit of Example 2 of the present invention. FIG. 11 is an explanatory diagram showing an example of payment judgment result information managed by a payment judgment result management unit according to a second embodiment of the present invention. 13 is a flowchart illustrating an example of a process executed by a payment determination unit according to the second embodiment of the present invention. 13 is an explanatory diagram showing an example of a user interface that displays billing information in the payment determination process by the information analysis support system according to the second embodiment of the present invention. FIG. 13 is an explanatory diagram illustrating an example of a user interface on which the information analysis support system according to the second embodiment of the present invention displays the results of the payment determination process. FIG.

The following describes an embodiment of the present invention with reference to the drawings.

Figure 1 is a block diagram showing an example of the configuration of an information analysis support system 101 according to a first embodiment of the present invention.

The information analysis support system 101 is a computer system, and includes an input unit 102, such as a keyboard and a mouse, an output unit 103 representing a display that outputs display data, a CPU (Central Processing Unit) 104, a memory 105, a communication unit 108, and a storage medium 106.

The information analysis support system 101 has a disease name information extraction unit 111, a drug information extraction unit 112, a test result information extraction unit 113, a surgery information extraction unit 114, a hospitalization information extraction unit 115, an explanatory variable information generation unit 116, a risk calculation unit 117, and a risk assessment unit 118. The functions of each unit from the disease name information extraction unit 111 to the risk assessment unit 118 are realized by the CPU 104 executing a program stored in the storage medium 106. When these programs are executed by the CPU 104, at least a portion of them may be copied to the memory 105 as necessary.

A database 107 is connected to the information analysis support system 101. The database 107 has a basic information management section 121, a health information management section 122, a disease name definition information management section 123, a drug definition information management section 124, a surgery definition information management section 125, a hospitalization definition information management section 126, a test result definition information management section 127, an explanatory variable definition information management section 128, an explanatory variable generation result management section 129, a risk calculation target information management section 130, a risk model information management section 131, and a classification definition information management section 132.

As described below, the basic information management unit 121 manages basic information 200 (Figure 2). The health information management unit 122 manages medical history information 300 (Figure 3) and medical checkup information 400 (Figure 4). The disease name definition information management unit 123 manages disease name definition information 500 (Figure 5). The drug definition information management unit 124 manages drug definition information 600 (Figure 6). The surgery definition information management unit 125 manages surgery definition information 700 (Figure 7). The hospitalization definition information management unit 126 manages hospitalization definition information 800 (Figure 8). The test result definition information management unit 127 manages test result definition information 900 (Figure 9). The explanatory variable definition information management unit 128 manages explanatory variable definition information 1000 (Figure 10). The explanatory variable generation result management unit 129 manages the intermediate table 1100 (Figure 11) and explanatory variable generation result 1300 (Figure 13) of the explanatory variable generation process. The risk calculation target information management unit 130 manages the risk calculation target information 1200 (Figure 12). The risk model information management unit 131 manages the risk model information 1400 (Figure 14). The classification definition information management unit 132 manages the classification definition information 1500 (Figure 15).

The database 107 may be stored, for example, in a storage system connected to the information analysis support system 101 via a network, or may be built into the information analysis support system 101 (for example, by being stored in the storage medium 106). When the database 107 is stored in a system external to the information analysis support system 101, at least a portion of its contents may be copied to the storage medium 106 or memory 105 as necessary. In addition, the entire system including the computer having the input unit 102, output unit 103, CPU 104, memory 105, and storage medium 106, and the database 107 may be referred to as the information analysis support system.

In addition, the information analysis support system 101 may be realized by a single computer having the configuration shown in FIG. 1, for example, or may be realized by multiple computers. For example, the information held in the database 107 described above may be stored in a distributed manner in multiple storage media 106 or memory 105, and the functions of the information analysis support system 101 described above may be executed in a distributed manner by multiple CPUs 104 of multiple computers.

Figure 2 is an explanatory diagram showing an example of basic information 200 managed by the basic information management unit 121 in Example 1 of the present invention.

Basic information 200 is basic information about each person who applies to subscribe to an insurance product from an insurance company.

Specifically, the basic information 200 includes a personal ID 201 that identifies each person, a gender 202 that identifies each person's gender, a date of birth 203 that identifies each person's date of birth, and an application acceptance date 204 that identifies the date on which an application from each person was accepted. The above basic information 200 is an example, and the basic information 200 may include various information regarding each person as necessary.

Figure 3 is an explanatory diagram showing an example of medical history information 300 managed by the health information management unit 122 in Example 1 of the present invention.

Medical condition information 300 is information about illnesses or injuries (medical conditions) experienced by each person in the past, extracted from the notification information submitted by each person to the insurance company when applying for insurance. For example, the information entered by each person in the notification information in the field corresponding to a medical condition may be retained as is as medical condition information 300. Medical condition information 300 includes personal ID 301, serial number 302, disease name 303, medication name 304, date of admission 305, date of discharge 306, surgery 307, etc.

The personal ID 301 is information that identifies each person, and corresponds to the personal ID 201 in the basic information 200. The sequence number 302 is information that identifies each of the multiple items of content extracted from the notification information of each person. The disease name 303, drug name 304, date of admission 305, date of discharge 306, and surgery 307 respectively indicate the content extracted from the disease name, drug name, date of admission, date of discharge, and surgery items of the notification information of each person.

Since the information entered by each person is registered as is for disease name 303 through surgery 307, it may contain inaccurate information, ambiguous information, or missing information. For example, disease name 303 may contain ambiguous information such as "I was told I have dyslipidemia." Alternatively, disease name 303 may be unknown (for example, the person does not remember it), but information on drug name 304 or surgery 307 may be available.

In the example of FIG. 3, the first and second lines of medical history information 300 indicate that a person identified by ID "P001" was hospitalized from April 1, 2020 to April 5, 2020 due to high blood pressure, and submitted notification information stating that a doctor diagnosed him with dyslipidemia. The third line indicates that a person identified by ID "P002" was prescribed drug A for an unknown illness, and underwent surgery A.

Figure 4 is an explanatory diagram showing an example of health check information 400 managed by the health information management unit 122 in Example 1 of the present invention.

The medical checkup information 400 is information about the results of a medical checkup (medical examination) taken by each person, and may be information submitted by each person as part of the notification information when applying for insurance. The medical checkup information 400 includes a personal ID 401, examination date 402, BMI 404, fasting blood glucose 404, HbA1c 405, interview results 406, and findings 407, etc.

The personal ID 401 is information that identifies each person, and corresponds to the personal ID 201 in the basic information 200. The examination date 402 indicates the date on which the health check was conducted. The BMI 404, fasting blood glucose 404, and HbA1c 405 are examples of test values obtained as a result of the health check. The medical interview result 406 is the result of the medical interview at the time of the health check, and may include information indicating, for example, whether or not the person has a drinking habit, whether or not they have an exercise habit, etc. The issues raised 407 are information on issues raised as a result of the health check.

The above information is a representative example of information obtained as a result of a health check, and in reality, the health check information 400 may not include at least any of these, or may include information on items other than these.

Figure 5 is an explanatory diagram showing an example of disease name definition information 500 managed by the disease name definition information management unit 123 in Example 1 of the present invention.

The information on disease names extracted from the notification information is aggregated into a number of predetermined categories according to the disease name definition information 500. That is, the disease name definition information 500 is dictionary information that associates descriptions of disease names included in the medical history information 300 with categories in which disease names are aggregated, and includes a disease name aggregation category 501, a target disease name 502, and an evaluation processing category (disease name) 503.

Disease name aggregation category 501 is information that identifies the category in which disease names are aggregated, and may be, for example, a standardized code such as the ICD-10 (International Classification of Diseases) code, or a code used independently by the insurance company. This disease name aggregation category 501 is associated with explanatory variables that are input to a risk model for evaluating the risk of insurance payment in the process described below.

The target disease name 502 is information that is expected to be extracted from the disease name field of the medical history information 300. As with the medical history information 300, the portion of the notification information filled in by the insurance applicant is not necessarily accurate information. For this reason, the information extracted from the disease name field of the medical history information 300 may also be inaccurate. For this reason, in order to correctly match not only accurate disease names but also vague or inaccurate disease name descriptions with categories that consolidate disease names, a description of the disease name expected to be entered in the notification information is registered in the target disease name 502.

In the example of FIG. 5, "Disease A", "Disease B", and "Disease C" are registered as target disease names 502 corresponding to the value "D1" of disease name aggregation category 501. For example, if "D1" is a code for identifying dyslipidemia, "Disease A" is the accurate disease name "Dyslipidemia", while "Disease B" and "Disease C" may be ambiguous terms such as "Didn't know I had dyslipidemia" and "High lipids", respectively.

Note that the description of the disease name included in the notification information may contain typographical errors (e.g., typos, etc.) because it is filled in by the person seeking insurance. For this reason, disease names containing anticipated typos (e.g., typos that have occurred frequently in the past) may be registered as target disease names 502. However, in the process of extracting information related to disease names from the notification information described below, typos may be able to be corrected by syntactic analysis. For this reason, the amount of data held may be reduced by not registering typos that can be corrected by syntactic analysis in target disease names 502.

The evaluation processing category (disease name) 503 is information indicating whether or not the value of the disease name aggregation category 501 identified according to the disease name definition information 500 is to be subject to risk evaluation by the risk model. The value "Target" of the evaluation processing category (disease name) 503 indicates that it is to be subject to risk evaluation processing by the model, "Not Target" indicates that it is not to be subject, and "Target (confirmation required)" indicates that it is to be subject but manual confirmation is required.

For example, if the disease name aggregation category "D1" that identifies "dyslipidemia" is identified based on the statement "I was told I have dyslipidemia," and the accuracy of the identification is sufficiently high, and a certain level of accuracy can be expected when a risk assessment is performed using a risk model based on the identified disease name aggregation category, then "target" may be registered as the evaluation processing category (disease name) 503.

On the other hand, for example, if the value of the target disease name 502 is only "lipids" and is presumed to indicate "dyslipidemia," but the certainty is low and a certain level of accuracy cannot be expected when a risk assessment is performed using a risk model based on the identified disease name aggregation category, "not target" may be registered as the evaluation processing category (disease name) 503. In this case, the notification information of a person who falls into that disease name aggregation category is excluded from the risk assessment using the risk model.

In addition, in cases where the accuracy of identifying the disease name aggregation category is moderate and a certain level of precision can be expected in the risk assessment if the identification result is correct, "Target (verification required)" may be registered as the evaluation processing category (disease name) 503. In this case, the notification information of a person who falls into the disease name aggregation category is subject to risk assessment using the risk model, but manual confirmation of the results is required. If it is manually confirmed that the identified disease name aggregation category is correct, the results of the risk assessment using the risk model may be adopted, and if not, the results of the risk assessment using the risk model may be rejected.

Figure 6 is an explanatory diagram showing an example of drug definition information 600 managed by the drug definition information management unit 124 in Example 1 of the present invention.

The information on drugs extracted from the notification information is aggregated into a number of predetermined categories in accordance with the drug definition information 600. That is, the drug definition information 600 is dictionary information that associates descriptions of drugs included in the medical history information 300 with categories in which the drugs are aggregated, and includes a drug aggregation category 601, a drug name 602, and an evaluation processing category (drug) 603.

The drug aggregation category 601 is information that identifies the category in which drugs are aggregated, and may be a standardized code, or a code that is used independently by the insurance carrier. This drug aggregation category 601 is associated with explanatory variables that are input to a risk model for evaluating the risk of insurance payment in the process described below.

The drug name 602 is information that is expected to be extracted from the drug item of the medical history information 300. As in the case of the disease name, in order to correctly match the drug name, which may be ambiguous or inaccurate, described in the notification information with the classification into which the drug is aggregated, a description of the drug name expected in the notification information is registered in the drug name 602.

The evaluation processing category (drug) 603 is information indicating whether or not the value of the drug aggregation category 601 identified according to the drug definition information 600 is to be subject to risk evaluation using a risk model. This information is similar to the evaluation processing category (disease name) 503 of the disease name definition information 500, so a detailed explanation is omitted.

Figure 7 is an explanatory diagram showing an example of surgery definition information 700 managed by the surgery definition information management unit 125 in Example 1 of the present invention.

Information about the surgery extracted from the notification information is aggregated into a number of predetermined categories in accordance with the surgery definition information 700. In other words, the surgery definition information 700 is dictionary information that associates the description of the surgery included in the medical history information 300 with the category in which the surgery is aggregated, and includes a surgery aggregation category 701, a surgery name 702, and an evaluation processing category (surgery) 703.

The surgery aggregation category 701 is information that identifies the category in which surgeries are aggregated, and may be a standardized code, or a code that is used independently by the insurance company. This surgery aggregation category 701 is associated with explanatory variables that are input to a risk model for evaluating the risk of insurance payment in the process described below.

The surgery name 702 is information that is expected to be extracted from the surgery item of the medical history information 300. As with the disease name, in order to correctly match the surgery name, which may be ambiguous or inaccurate, described in the notification information with the category that consolidates the surgery names, a description of the surgery name expected in the notification information is registered in the surgery name 702.

The evaluation processing category (surgery) 703 is information indicating whether or not the value of the surgery aggregation category 701 identified according to the surgery definition information 700 is to be subject to risk evaluation using a risk model. This information is similar to the evaluation processing category (disease name) 503 of the disease name definition information 500, so a detailed explanation is omitted.

Figure 8 is an explanatory diagram showing an example of hospitalization definition information 800 managed by the hospitalization definition information management unit 126 in Example 1 of the present invention.

Information about hospitalization extracted from the notification information is aggregated into a number of predetermined categories in accordance with the hospitalization definition information 800. In other words, the hospitalization definition information 800 is dictionary information that associates the contents of hospitalization included in the medical history information 300 with the categories in which hospitalizations are aggregated, and includes a hospitalization aggregation category 801, number of days of hospitalization 802, and an evaluation processing category (hospitalization) 803.

The hospitalization aggregation category 801 is information that identifies the category in which hospitalizations are aggregated, and may be a standardized code, or a code that is used independently by the insurance company. This hospitalization aggregation category 801 is associated with explanatory variables that are input to a risk model for evaluating the risk of insurance payment in the process described below.

The number of days of hospitalization 802 is information on the number of days of hospitalization that is expected to be extracted from the surgery item of the medical history information 300. For example, the number of days of hospitalization 802 may be information indicating a range of the number of days of hospitalization.

The evaluation processing category (hospitalization) 803 is information indicating whether the value of the hospitalization aggregation category 801 identified according to the hospitalization definition information 800 is to be subject to risk evaluation using a risk model.

In this embodiment, information about hospitalizations is aggregated into multiple categories according to the number of days of hospitalization. In the example of Figure 8, information about the number of days of hospitalization "0" (i.e., not hospitalized) is aggregated into hospitalization aggregation category "A1." Information about the number of days of hospitalization "1" to "14" is aggregated into hospitalization aggregation category "A2." Information about the number of days of hospitalization "15" to "31" is aggregated into hospitalization aggregation category "A3." Information about the number of days of hospitalization "32" or more is aggregated into hospitalization aggregation category "A4."

Furthermore, in the example of FIG. 8, the evaluation processing classifications of the hospitalization aggregation classifications "A1" and "A2" are "target", the evaluation processing classification of the hospitalization aggregation classification "A3" is "target (confirmation required)", and the evaluation processing classification of the hospitalization aggregation classification "A4" is "non-target". In general, the number of samples of notification information included in the teacher data used to construct a risk model tends to decrease as the number of days of hospitalization increases. Therefore, the more days of hospitalization in the notification information input as an explanatory variable to the constructed risk model, the lower the frequency of occurrence of the number of samples of such notification information, and the accuracy of the risk value calculated by the risk model tends to decrease. Therefore, in the above example, the hospitalization aggregation classification "A3" is included in the target of risk calculation using the risk model, but manual confirmation is required, and the hospitalization aggregation classification "A4" is excluded from the target of risk calculation using the risk model.

Figure 9 is an explanatory diagram showing an example of test result definition information 900 managed by the test result definition information management unit 127 in Example 1 of the present invention.

Information such as test results extracted from the medical checkup information 400 submitted at the time of notification is aggregated into a number of predetermined categories in accordance with the test result definition information 900. In other words, the test result definition information 900 is dictionary information that associates the test results included in the medical checkup information 400 with the category in which the test results are aggregated, and includes a test result aggregation category 901, test results 902, and an evaluation processing category (test) 903.

The test result aggregation category 901 is information that identifies the category in which the test results are aggregated, and may be a standardized code, or a code that is used independently by the insurance company. This test result aggregation category 901 is associated with explanatory variables that are input to a risk model for evaluating the risk of insurance payment in the process described below.

The test result 902 is information on the test result that is expected to be extracted from the health check information 400. For example, the test result 902 may be information that specifies a range of values of the test result.

The evaluation processing category (test) 903 is information indicating whether or not the value of the test result aggregation category 901 identified according to the test result definition information 900 is to be subject to risk evaluation by the risk model. As in the case of the evaluation processing category (hospitalization) 803, for a test result aggregation category corresponding to a range of test result values that appear infrequently in the data used to construct the risk model, "subject (confirmation required)" or "not subject" may be registered as the evaluation processing category (test) 903.

Figure 10 is an explanatory diagram showing an example of explanatory variable definition information 1000 managed by the explanatory variable definition information management unit 128 in Example 1 of the present invention.

Explanatory variable definition information 1000 is dictionary information that associates explanatory variables with combinations of aggregation results for each item category based on each person's notification information. Specifically, explanatory variable definition information 1000 includes explanatory variables 1001, disease name aggregation category 1002, medication aggregation category 1003, test aggregation category 1004, surgery aggregation category 1005, and hospitalization aggregation category 1006. This associates the value of explanatory variable 1001 with combinations of values from disease name aggregation category 1002 to hospitalization aggregation category 1006.

For example, if the disease name aggregation category based on a person's notification information is "D1" and there are no corresponding values from the medication aggregation category to the hospitalization aggregation category, the explanatory variable corresponding to that person's notification information will be "E101." Also, if the disease name aggregation category based on another person's notification information is "D1," there are no corresponding values from the medication aggregation category to the surgery aggregation category, and the hospitalization aggregation category is either "A1," "A2," or "A3," the explanatory variable corresponding to that person's notification information will be "E102."

Figure 11 is an explanatory diagram showing an example of an intermediate table 1100 of the explanatory variable generation process managed by the explanatory variable generation result management unit 129 in Example 1 of the present invention.

The intermediate table 1100 of the explanatory variable generation process is information that is compiled by aggregating information extracted from each person's notification information based on disease name definition information 500, drug definition information 600, surgery definition information 700, hospitalization definition information 800, and test result definition information 900, and is generated during the explanatory variable generation process.

Specifically, the intermediate table 1100 of the explanatory variable generation process includes an individual ID 1101, a disease name aggregation category 1102, a medication aggregation category 1103, a test aggregation category 1104, a surgery aggregation category 1105, a hospitalization aggregation category 1106, an evaluation processing category (disease name) 1107, an evaluation processing category (medication) 1108, an evaluation processing category (examination) 1109, an evaluation processing category (surgery) 1110, and an evaluation processing category (hospitalization) 1111.

The individual ID 1101 is information that identifies each person, and corresponds to the individual ID 201 in the basic information 200. The disease name aggregation category 1102, medication aggregation category 1103, test aggregation category 1104, surgery aggregation category 1105, hospitalization aggregation category 1106, evaluation processing category (disease name) 1107, evaluation processing category (medication) 1108, evaluation processing category (test) 1109, evaluation processing category (surgery) 1110, and evaluation processing category (hospitalization) 1111 indicate the results of aggregating information for each item of each person's notification information and the evaluation processing category for each item.

Figure 12 is an explanatory diagram showing an example of risk calculation target information 1200 managed by the risk calculation target information management unit 130 in Example 1 of the present invention.

The risk calculation target information 1200 includes information for determining whether or not each person's notification information is to be subject to evaluation by the risk model, based on the value of the evaluation processing category for each item of the notification information for each person. Specifically, the risk calculation target information 1200 includes evaluation target category 1201, evaluation processing category (disease name) 1202, evaluation processing category (medication) 1203, evaluation processing category (examination) 1204, and evaluation processing category (surgery) 1205.

Evaluation processing category (disease name) 1202, evaluation processing category (medication) 1203, evaluation processing category (examination) 1204, and evaluation processing category (surgery) 1205 are evaluation processing categories extracted from the notification information of each person, and correspond to evaluation processing category (disease name) 503, evaluation processing category (medication) 603, evaluation processing category (examination) 903, and evaluation processing category (surgery) 703, respectively. Evaluation target category 1201 is the final evaluation target category of the notification information of each person (i.e., information indicating whether or not the notification information of each person is to be evaluated by the risk model) corresponding to the set of values of evaluation processing category (disease name) 1202 to evaluation processing category (surgery) 1205. The value of evaluation target category 1201 is one of "target", "target (confirmation required)", and "not target", just like the evaluation processing category of individual items.

For example, if the evaluation processing category of all items extracted from a person's notification information is "target," the notification information of that person may be determined to be subject to evaluation by the risk model. Also, if the evaluation processing category of at least one item extracted from a person's notification information is "target (requires confirmation)" and the evaluation processing category of all remaining items is "target," it may be determined that the notification information of that person is subject to evaluation by the risk model but requires manual confirmation. Also, if the evaluation processing category of at least one item extracted from a person's notification information is "not target," it may be determined that the notification information of that person is excluded from the target of evaluation by the risk model.

However, the above-mentioned criteria are merely examples, and other criteria may be applied. In addition, in this embodiment, the three-level judgment, "eligible," "eligible (confirmation required)," and "non-eligible," is performed as described above, but it is also possible to perform judgment using only two levels, for example, "eligible" and "eligible (confirmation required)," or only two levels, "eligible" and "non-eligible."

Figure 13 is an explanatory diagram showing an example of explanatory variable generation result information 1300 managed by the explanatory variable generation result management unit 129 in Example 1 of the present invention.

The explanatory variable generation result information 1300 includes information on explanatory variables generated as a result of the explanatory variable generation process. Specifically, the explanatory variable generation result information 1300 includes personal ID 1301, gender 1302, age 1303, BMI 1304, fasting blood glucose 1305, HbA1c 1306, interview result 1307, findings 1308, E101_1309, E102_1310, E103_1311, E201_1312, E202_1313, E203_1314, and evaluation target category 1315.

The individual ID 1301 is information that identifies each person, and corresponds to the individual ID 201 in the basic information 200. The gender 1302 and age 1303 correspond to the information of each person extracted from the basic information 200. The BMI 1304 to the findings 1308 correspond to the information of each person extracted from the medical checkup information 400. FIG. 13 is an example, and the explanatory variable generation result information 1300 includes necessary items depending on the items of the basic information 200 and the medical checkup information 400.

E101_1309 to E203_1314 indicate explanatory variables identified by comparing the information extracted from each person's notification information with the explanatory variable definition information 1000 from the disease name definition information 500. For example, a "1" is registered in the field corresponding to the identified explanatory variable. Although only some explanatory variables are shown in FIG. 13, in reality all of the identified explanatory variables are registered.

The evaluation target category 1315 indicates the evaluation target category identified by comparing the information extracted from each person's notification information with the test result definition information 900 and the risk calculation target information 1200 from the disease name definition information 500.

Note that in reality, a single person may submit notification information for multiple diseases, multiple medications, etc. For this reason, multiple explanatory variables may be identified from a single person's notification information. In that case, a "1" is registered in the corresponding explanatory variable column among E101_1309 to E203_1314, etc. Also, a single explanatory variable may be identified multiple times from a single person's notification information. In that case, a "1" is registered in the column for that explanatory variable.

Figure 14 is an explanatory diagram showing an example of risk model information 1400 managed by the risk model information management unit 131 in Example 1 of the present invention.

Risk model information 1400 includes a model ID 1401 that identifies the model (risk model) that assesses risk, a model name 1402 that indicates the type of model each model is, and model parameters 1403 that indicate the structure and parameters of the model.

The risk model information 1400 can register parameters for multiple risk models, and any one of them can be selected as needed and used to calculate risk.

Figure 15 is an explanatory diagram showing an example of classification definition information 1500 managed by the classification definition information management unit 132 in Example 1 of the present invention.

Classification definition information 1500 is information that associates a risk value calculated by a risk model with a classification that asks whether or not insurance is possible, and includes model ID 1501, classification conditions 1502, and classification label 1503. Model ID 1501 is information that identifies a risk model, and corresponds to model ID 1401 of risk model information 1400. Classification conditions 1502 are conditions that are applied to the risk value calculated by the risk model, and may be, for example, information indicating a range of risk values. Classification label 1503 is information that indicates a classification, such as whether or not a person who satisfies the conditions indicated by classification conditions 1502 is able to subscribe to insurance.

In the example of Figure 10, if the risk value calculated by applying the risk model identified by model ID "1" to the notification information is greater than or equal to 0 and less than 0.7, the person who submitted the notification information is classified as a standard person (i.e., a person who can take out insurance unconditionally), if the risk value is greater than or equal to 0.7 and less than 0.9, the person is classified as a conditional person (i.e., a person who can take out insurance under certain conditions, such as an increase in health insurance premiums), and if the risk value is greater than or equal to 0.9, the person is classified as a rejected person (i.e., a person who cannot take out insurance).

The example in Figure 10 shows the definition of a classification corresponding to model ID "1", but if the risk model information 1400 holds multiple risk models, the classification definition information 1500 holds information that defines the classification for each risk model. This allows appropriate classification for each risk model.

Figure 16 is a flowchart showing an example of processing executed by the disease name information extraction unit 111 in Example 1 of the present invention.

When processing starts (step 1601), the disease name information extraction unit 111 reads the notification information (step 1602). Next, the disease name information extraction unit 111 reads the disease name definition information 500 (step 1603).

Next, the disease name information extraction unit 111 extracts disease name information from the read notification information (step 1604). This extracts, for example, the value of disease name 303 from the medical history information 300. Note that the information on disease name included in the notification information may contain typographical errors because it was entered by the person wishing to enroll in insurance. For this reason, the disease name information extraction unit 111 may perform syntax analysis to correct any typographical errors included in the information on disease name.

Next, the disease name information extraction unit 111 identifies a disease name aggregation category based on the extracted disease name information and the disease name definition information 500 (step 1605). This identifies, for example, the value of the disease name aggregation category 501 of the disease name definition information 500 that corresponds to the value of the extracted disease name 303.

Next, the disease name information extraction unit 111 stores the identified disease name aggregation category (step 1606). As a result, for example, the value of the identified disease name aggregation category 501 and the value of the corresponding evaluation processing category (disease name) 503 are stored in the disease name aggregation category 1102 and evaluation processing category (disease name) 1107 of the intermediate table 1100 of the explanatory variable generation process, respectively.

This completes the processing of the disease name information extraction unit 111 (step 1607).

Figure 17 is a flowchart showing an example of processing executed by the drug information extraction unit 112 in Example 1 of the present invention.

When processing starts (step 1701), the drug information extraction unit 112 reads notification information (step 1702). Next, the drug information extraction unit 112 reads the drug definition information 600 (step 1703).

Next, the drug information extraction unit 112 extracts drug information from the read notification information (step 1704). This extracts, for example, the value of the drug name 304 from the medical history information 300. As in the case of disease name information, the drug information extraction unit 112 may perform syntax analysis to correct typographical errors contained in the information about the drug.

Next, the drug information extraction unit 112 identifies a drug aggregation category based on the extracted drug information and the drug definition information 600 (step 1705). This identifies, for example, the value of the drug aggregation category 601 in the drug definition information 600 that corresponds to the value of the extracted drug name 304.

Next, the drug information extraction unit 112 stores the identified drug aggregation category (step 1706). As a result, for example, the value of the identified drug aggregation category 601 and the value of the corresponding evaluation processing category (drug) 603 are stored in the drug aggregation category 1103 and evaluation processing category (drug) 1108 of the intermediate table 1100 of the explanatory variable generation process, respectively.

This completes the processing of the drug information extraction unit 112 (step 1707).

Figure 18 is a flowchart showing an example of processing executed by the test result information extraction unit 113 in Example 1 of the present invention.

When processing starts (step 1801), the test result information extraction unit 113 reads the notification information (step 1802). Next, the test result information extraction unit 113 reads the test result definition information 900 (step 1803).

Next, the test result information extraction unit 113 extracts the test result information from the read notification information (step 1804). This extracts, for example, information corresponding to the test results contained in the medical examination information 400.

Next, the test result information extraction unit 113 identifies the test result aggregation category based on the extracted test result information and the test result definition information 900 (step 1805). This identifies, for example, the value of the test result aggregation category 901 of the test result definition information 900 that corresponds to the extracted test result.

Next, the test result information extraction unit 113 stores the identified test result aggregation category (step 1806). As a result, for example, the value of the identified test result aggregation category 901 and the value of the corresponding evaluation processing category (test) 903 are stored in the test aggregation category 1104 and evaluation processing category (test) 1109 of the intermediate table 1100 of the explanatory variable generation process, respectively.

This completes the processing of the test result information extraction unit 113 (step 1807).

Figure 19 is a flowchart showing an example of processing executed by the surgery information extraction unit 114 in Example 1 of the present invention.

When processing starts (step 1901), the surgery information extraction unit 114 reads the notification information (step 1902). Next, the surgery information extraction unit 114 reads the surgery definition information 700 (step 1903).

Next, the surgery information extraction unit 114 extracts surgery information from the notification information that has been read (step 1904). This extracts, for example, the value of surgery 307 from the medical history information 300.

Next, the surgery information extraction unit 114 identifies the surgery aggregation category based on the extracted surgery information and the surgery definition information 700 (step 1905). This identifies, for example, the value of the surgery aggregation category 701 in the surgery definition information 700 that corresponds to the value of the extracted surgery 307.

Next, the surgery information extraction unit 114 stores the identified surgery aggregation category (step 1906). As a result, for example, the value of the identified surgery aggregation category 701 and the value of the corresponding evaluation processing category (surgery) 703 are stored in the surgery aggregation category 1105 and evaluation processing category (surgery) 1110, respectively, in the intermediate table 1100 of the explanatory variable generation process.

This completes the processing of the surgery information extraction unit 114 (step 1907).

Figure 20 is a flowchart showing an example of processing executed by the hospitalization information extraction unit 115 in Example 1 of the present invention.

When processing starts (step 2001), the hospitalization information extraction unit 115 reads the notification information (step 2002). Next, the hospitalization information extraction unit 115 reads the hospitalization definition information 800 (step 2003).

Next, the hospitalization information extraction unit 115 extracts the hospitalization information from the read notification information (step 2004). This extracts, for example, the values of the admission date 305 and the discharge date 306 from the medical history information 300.

Next, the hospitalization information extraction unit 115 identifies the hospitalization aggregation category based on the extracted hospitalization information and the hospitalization definition information 800 (step 2005). This identifies, for example, the value of the hospitalization aggregation category 801 in the hospitalization definition information 800 that corresponds to the number of days of hospitalization identified from the extracted values of the admission date 305 and discharge date 306.

Next, the hospitalization information extraction unit 115 stores the identified hospitalization aggregation category (step 2006). As a result, for example, the value of the identified hospitalization aggregation category 801 and the value of the corresponding evaluation processing category (hospitalization) 803 are stored in the hospitalization aggregation category 1106 and the evaluation processing category (hospitalization) 1111 of the intermediate table 1100 of the explanatory variable generation process, respectively.

This completes the processing of the hospitalization information extraction unit 115 (step 2007).

Figure 21 is a flowchart showing an example of processing executed by the explanatory variable information generation unit 116 in Example 1 of the present invention.

When the process starts (step 2101), the explanatory variable information generation unit 116 reads the aggregation classification results (step 2102). For example, the explanatory variable information generation unit 116 may read the intermediate table 1100 of the explanatory variable generation process.

Next, the explanatory variable information generation unit 116 identifies explanatory variables by matching the explanatory variable definition information 1000 with the intermediate table 1100 of the explanatory variable generation process that was read (step 2103).

Next, the explanatory variable information generation unit 116 groups the information extracted from the notification information and the explanatory variables obtained as a result of the matching by organizing them according to each person's ID (step 2104).

Next, the explanatory variable information generation unit 116 stores the grouped information as explanatory variable generation result information 1300 (step 2105).

This completes the processing of the explanatory variable information generation unit 116 (step 2105).

Figure 22 is a flowchart showing an example of processing executed by the risk calculation unit 117 in Example 1 of the present invention.

When the process starts (step 2201), the risk calculation unit 117 sets the conditions for the risk calculation process (step 2202). The conditions set here may include, for example, specification of notification information for which risk is to be calculated, the risk model to be used, data items to be aggregated, etc.

Next, the risk calculation unit 117 reads the explanatory variable information (step 2203). For example, the risk calculation unit 117 may read the explanatory variable generation result information 1300.

Next, the risk calculation unit 117 calculates the risk by inputting the explanatory variables included in the explanatory variable generation result information 1300 into the risk model (step 2204).

Next, the risk calculation unit 117 stores the calculated risk value in the storage medium 106 or the database 107 (step 2205). This ends the processing of the risk calculation unit 117 (step 2206).

Figure 23 is a flowchart showing an example of processing executed by the risk assessment unit 118 in Example 1 of the present invention.

When processing begins (step 2301), the risk assessment unit 118 reads the risk value calculated by the risk calculation unit 117 (step 2302).

Next, the risk assessment unit 118 classifies the risk calculation results based on each person's risk model into those that are subject to risk assessment (e.g., those that fall under "target" or "target (confirmation required)") and those that are not subject to risk assessment (e.g., those that fall under "not subject") based on the value of the evaluation target category 1315 for each person who is the target of risk assessment (step 2303).

Next, the risk assessment unit 118 adds a classification label to the risk calculation result of each person who is the subject of risk assessment by matching the classification definition information 1500 corresponding to the risk model used in the risk calculation (step 2304). Specifically, the value of the classification label 1503 corresponding to the value of the classification condition 1502 satisfied by the risk value calculated for each person is assigned.

Next, the risk assessment unit 118 stores the assigned classification label in the storage medium 106 or the database 107 (step 2305). At this time, if the risk assessment unit 118 determines that the item is a "target (confirmation required)" based on the risk calculation target information 1200, it may store information indicating this together with the classification label in the storage medium 106 or the database 107. In addition, the risk assessment unit 118 may output this information via the output unit 103 or the communication unit 108 (see FIG. 27).

This completes the processing of the risk assessment unit 118 (step 2306).

Next, an example of a user interface provided by the information analysis support system 101 when executing the processes shown in Figures 16 to 23 will be described with reference to Figures 24 to 27.

Figure 24 is an explanatory diagram showing an example of a user interface that displays notification information in the process of generating explanatory variables by the information analysis support system 101 according to the first embodiment of the present invention.

The explanatory variable generation screen 2400 shown in FIG. 24 is an example of display data output by the information analysis support system 101, and includes, for example, a notification information designation section 2401, a notification information read execution button 2402, an aggregation processing target designation section 2403, an aggregation processing execution button 2404, an explanatory variable generation processing execution button 2405, and an information display section 2406.

This display data may be output as an image by the output unit 103, or may be output by the communication unit 108. In the latter case, the display data may be transferred from the communication unit 108 to an external device (such as a terminal device used by a user, not shown) via a network (not shown) and output as an image by the external device. In this case, information input via the screen is input using an input unit (not shown) of the external device, and input to the information analysis support system 101 via the network and the communication unit 108. The same applies to the other screens described below.

Information specifying the notification information to be read is input to the notification information specification section 2401. When the user inputs information specifying the notification information to the notification information specification section 2401 and operates the notification information read execution button 2402, the specified notification information is read (see steps 1602, 1702, 1802, 1902 and 2002). The information display section 2406 displays the read notification information 2407. The notification information 2407 may be the same as, for example, the medical history information 300 and the medical checkup information 400.

Figure 25 is an explanatory diagram showing an example of a user interface that displays the results of the aggregation process in the process of generating explanatory variables by the information analysis support system 101 according to the first embodiment of the present invention.

For example, after notification information is read as shown in FIG. 24, when the user inputs information specifying the information items to be aggregated in the aggregation processing target specification section 2403 and operates the aggregation processing execution button 2404, the aggregation processing of the specified items is executed (see steps 1605, 1705, 1805, 1905, and 2005). The aggregation processing result 2501 is displayed in the information display section 2406. The aggregation processing result 2501 may be the same as, for example, the intermediate table 1100 of the explanatory variable generation process.

Figure 26 is an explanatory diagram showing an example of a user interface that displays the results of the explanatory variable generation process in the process in which the information analysis support system 101 of Example 1 of the present invention generates explanatory variables.

For example, notification information is read as shown in FIG. 24, and aggregation processing is executed as shown in FIG. 25. Then, when the user operates the explanatory variable generation processing execution button 2405, the explanatory variable generation processing is executed (see FIG. 21). The information display section 2406 displays the explanatory variable generation result 2601. The explanatory variable generation result 2601 may be the same as, for example, the explanatory variable generation result information 1300.

Figure 27 is an explanatory diagram showing an example of a user interface that displays the results of a risk analysis performed by the information analysis support system 101 in Example 1 of the present invention.

The risk analysis execution screen 2700 shown in FIG. 27 includes, for example, an explanatory variable designation section 2701, an explanatory variable load execution button 2702, a risk model designation section 2703, a classification condition designation section 2704, a risk analysis execution button 2705, and an information display section 2706.

Information specifying the explanatory variables to be read is input to the explanatory variable specification section 2701. When the user inputs information specifying explanatory variables to the explanatory variable specification section 2701 and operates the explanatory variable read execution button 2702, the specified explanatory variables are read (see step 2203 in FIG. 22). For example, the explanatory variable generation result information 1300 may be read.

The risk model designation unit 2703 and the classification condition designation unit 2704 are respectively input with information designating a risk model used to calculate a risk value and information designating classification conditions based on the risk value. The risk model and classification conditions are designated according to user operations on the risk model designation unit 2703 and the classification condition designation unit 2704. For example, one of the risk models registered in the risk model information 1400 and one of the classification definitions registered in the classification definition information 1500 may be designated.

When the user operates the execute risk analysis button 2705, a risk analysis is executed based on the specified explanatory variables, the specified risk model, and the specified classification conditions (see step 2204 onwards in FIG. 22 and FIG. 23). The information display section 2706 displays a risk analysis result 2707. The risk analysis result 2707 is stored in the storage medium 106 or the database 107 as a result of processing by the risk calculation section 117 and the risk determination section 118, and includes, for example, a personal ID 2707A, a risk value 2707B, a classification result 2707C, and confirmation required 2707D.

The personal ID 2707A is information that identifies each person, and corresponds to the personal ID 201 in the basic information 200. The risk value 2707B is a risk value calculated by the risk calculation unit 117 by inputting explanatory variables generated from the notification information of each person into a risk model. The classification result 2707C is the classification of each person determined by the risk determination unit 118 by matching the risk value calculated for each person with the classification definition information 1500.

Confirmation required 2707D indicates whether the evaluation target category of the explanatory variable generated from each person's notification information is "target (confirmation required)." A value of "confirmation required" for confirmation required 2707D indicates that the evaluation target category of the explanatory variable is "target (confirmation required)," and a blank value for confirmation required 2707D ("-") indicates that the evaluation target category of the explanatory variable is "target." Note that explanatory variables with an evaluation target category of "non-target" are excluded from the analysis by the risk model, and the results are not displayed as risk analysis results 2707.

The user can refer to the risk analysis result 2707 and use as is the values of the classification result 2707C for which the corresponding value of confirmation required 2707D is not "confirmation required", and can use the values of the classification result 2707C for which the corresponding value of confirmation required 2707D is "confirmation required" after confirming that the aggregation during the generation of the explanatory variables was appropriate.

The functions of the information analysis support system 101 of the first embodiment described above may be provided via an API (Application Programming Interface). For example, when the information analysis support system 101 receives notification information via a network (not shown) connected to the communication unit 108, it may store the notification information in the database 107, execute the processes shown in FIGS. 16 to 23, and output data required to display the information shown in FIGS. 24 to 27 via the network.

According to the above-described first embodiment, by appropriately aggregating the values of the notification information items, risk assessment based on notification information including a variety of expressions can be automated using a risk model. Furthermore, when notification information satisfies a predetermined condition, the information can be excluded from the automation of risk assessment using the risk model, or risk assessment using the risk model can be performed but manual confirmation is required, thereby ensuring the accuracy of risk assessment.

Next, a second embodiment of the present invention will be described. Except for the differences described below, each part of the system of the second embodiment has the same function as each part of the system of the first embodiment shown in Figures 1 to 27 and marked with the same reference numerals, and therefore the description of those parts will be omitted.

Figure 28 is a block diagram showing an example of the configuration of an information analysis support system 101 according to the second embodiment of the present invention.

In the information analysis support system 101 of Example 2, the storage medium 106 further includes a payment judgment unit 119. The database 107 further includes a contract information management unit 133, a billing information management unit 134, a payment classification definition information management unit 135, and a payment judgment result information management unit 136. The function of the payment judgment unit 119, like the functions of the other units, is realized by the CPU 104 executing a program stored in the storage medium 106.

Figure 29 is an explanatory diagram showing an example of contract information 2900 managed by the contract information management unit 133 in Example 2 of the present invention.

Contract information 2900 is basic information about the insurance products of the insurance companies with which each person has a contract.

Specifically, the contract information 2900 includes a personal ID 2901 that identifies each person who has subscribed to the insurance product, a gender 2902 that identifies the gender of each person, a contracted insurance product 2903 that identifies the insurance product that each person has subscribed to, a daily hospitalization amount 2904 that indicates the amount paid per day of hospitalization for each insurance product, and a surgery benefit rate 2905 that indicates the rate of payment if surgery is performed. The above contract information 2900 is an example, and the contract information 2900 may include various information regarding the insurance products that each person has subscribed to as necessary.

Figure 30 is an explanatory diagram showing an example of billing information 3000 managed by the billing information management unit 134 in Example 2 of the present invention.

The claim information 3000 is information extracted from a claim for insurance payment made by an insurance policyholder to an insurance company, and for example, the contents written by each person in each item of the claim document may be retained as is as the claim information 3000. The claim information 3000 includes an individual ID 3001, a claim ID 3002, a claim date and time 3003, claim details 3004, number of days of hospitalization 3005, and surgery 3006.

The individual ID 3001 is information that identifies each person, and corresponds to the individual ID 2901 in the contract information 2900. The billing ID 3002 is information that identifies a bill from each person. The billing date and time 3003 is information that identifies the date and time when a bill was made by each person.

Claim details 3004 indicates the details of the claim from each person. For example, the details written by each person in the document for claim may be registered as claim details 3004 as is. Specifically, claim details 3004 may include the event (e.g., the name of the illness) that was the reason for claiming payment of insurance money.

Number of days hospitalized 3005 indicates the number of days each person was hospitalized, if any. Surgery 3006 indicates whether each person underwent surgery and, if so, what type of surgery they underwent.

Although omitted in FIG. 30, the billing information 3000 may include information indicating the medications prescribed for each person and information indicating each person's test results.

Figure 31 is an explanatory diagram showing an example of payment category definition information 3100 managed by the payment category definition information management unit 135 in Example 2 of the present invention.

The payment category definition information 3100 is dictionary information that associates the combination of aggregation results for each item category based on each person's billing information 3000 with a payment category. Specifically, the payment category definition information 3100 includes a payment category 3101, a disease name aggregation category 3102, a medication aggregation category 3103, an examination aggregation category 3104, a surgery aggregation category 3105, and a hospitalization aggregation category 3106. This associates the value of the payment category 3101 with a combination of values from the disease name aggregation category 3102 to the hospitalization aggregation category 3106.

For example, if the disease name aggregation category based on a person's billing information 3000 is "D1" and there are no corresponding values from the medication aggregation category to the hospitalization aggregation category, the payment category corresponding to the person's billing information 3000 will be "P101." The amount of insurance payment can be determined based on the identified payment category.

Figure 32 is an explanatory diagram showing an example of payment judgment result information 3200 managed by the payment judgment result information management unit 136 in Example 2 of the present invention.

The payment determination result information 3200 includes a personal ID 3201, a payment category 3202, a paid insurance amount 3203, an evaluation target category 3204, a disease name aggregation category 3205, a drug aggregation category 3206, an examination aggregation category 3207, a surgery aggregation category 3208, and a hospitalization aggregation category 3209.

The individual ID 3201 is information that identifies each person, and corresponds to the individual ID 2901 in the contract information 2900. The payment category 3202 indicates the payment category determined by the processing of the payment determination unit 119, which will be described later. The insurance payment amount 3203 indicates the amount of insurance payment determined according to the determined payment category. The evaluation target category 3204 indicates the evaluation target category determined by the processing of the payment determination unit 119, which will be described later. This is determined in the same way as the evaluation target category in Example 1.

Disease name aggregation category 3205 to hospitalization aggregation category 3209 each indicate the result of aggregating information on disease names, medications, test values, surgery, and hospitalizations extracted from contract information 2900 into a specified category. Aggregation into these categories is performed in the same manner as in Example 1. The value of payment category 3101 corresponding to the combination of disease name aggregation category 3102 to hospitalization aggregation category 3106 that matches the combination of values of disease name aggregation category 3205 to hospitalization aggregation category 3209 is registered as payment category 3202.

Figure 33 is a flowchart showing an example of processing executed by the payment determination unit 119 in Example 2 of the present invention.

When processing begins (step 3301), the payment determination unit 119 reads the billing information 3000 (step 3302). Next, the payment determination unit 119 sequentially extracts disease name information, drug information, test result information, hospitalization information, and surgery information from the billing information 3000 (steps 3303 to 3307).

Next, the payment determination unit 119 performs a payment determination for the extracted information (step 3308). Specifically, the payment determination unit 119 first performs a process of consolidating the extracted disease name information, drug information, test result information, surgery information, and hospitalization information into a predetermined category. This is the same as the process performed by the disease name information extraction unit 111, drug information extraction unit 112, test result information extraction unit 113, surgery information extraction unit 114, and hospitalization information extraction unit 115 in Example 1 (see steps 1605, 1705, 1805, 1905, and 2002). The categories consolidated in this way are registered in the disease name consolidation category 3205, drug consolidation category 3206, test consolidation category 3207, surgery consolidation category 3208, and hospitalization consolidation category 3209 of the payment determination result information 3200.

Then, the payment determination unit 119 identifies a combination of the disease name summary category 3102 to the hospitalization summary category 3106 that matches the combination of values of the disease name summary category 3205 to the hospitalization summary category 3209, and identifies the value of the payment category 3101 that corresponds to the identified combination as the payment determination result.

Next, the payment judgment unit 119 stores the identified payment judgment result in the payment category 3202 of the payment judgment result information 3200 (step 3309). This completes the processing of the payment judgment unit 119 (step 3310).

Next, an example of a user interface provided by the information analysis support system 101 of the second embodiment will be described with reference to Figures 34 and 35.

Figure 34 is an explanatory diagram showing an example of a user interface that displays billing information in the payment determination process by the information analysis support system 101 of Example 2 of the present invention.

The payment judgment screen 3400 shown in FIG. 34 is an example of display data output by the information analysis support system 101, and includes, for example, a billing information designation section 3401, a billing information read execution button 3402, an aggregation processing target designation section 3403, an aggregation processing execution button 3404, a payment judgment processing execution button 3405, and an information display section 3406.

In the billing information specification section 3401, information specifying the billing information to be read is input. When the user inputs information specifying the billing information in the billing information specification section 3401 and operates the billing information read execution button 3402, the specified billing information is read (see step 3302). The information display section 3406 displays the read billing information 3407. The billing information 3407 may be the same as the billing information 3000, for example.

Figure 35 is an explanatory diagram showing an example of a user interface in which the information analysis support system 101 of the second embodiment of the present invention displays the results of the payment determination process.

For example, as shown in FIG. 34, after the billing information is read, when the user inputs information specifying the information items to be aggregated in the aggregation processing target specification section 3403 and operates the aggregation processing execution button 3404, the aggregation processing of the specified items is executed in the payment determination process (step 3308).

Furthermore, when the user operates the payment judgment process execution button 3405, the process of identifying the payment judgment result by matching the aggregated category combinations is executed as part of the payment judgment process (step 3308). The judgment process result 3501 is displayed in the information display section 3406. The judgment process result 3501 may be the same as, for example, the payment judgment result information 3200.

The functions of the information analysis support system 101 in the above-described second embodiment may be provided via an API (Application Programming Interface) in the same manner as in the first embodiment. For example, when the information analysis support system 101 receives billing information via a network connected to the communication unit 108, it may store the billing information in the database 107, execute the process shown in FIG. 33, and output data required to display the information shown in FIGS. 34 to 35 via the network.

The information analysis support system 101 shown in FIG. 28 has a risk analysis function similar to that of Example 1, in addition to the function of identifying payment categories based on the billing information described above. However, this system configuration is only an example, and the information analysis support system 101 of Example 2 does not need to have the risk analysis function as in Example 1.

According to the above-mentioned Example 2, the insurance payment determination can be automated while ensuring accuracy.

The system of the present invention may also be configured as follows:

(1) An information analysis support method executed by a computer system (e.g., information analysis support system 101), the computer system having a processor (e.g., CPU 104) and a storage device (e.g., storage medium 106 or database 107) connected to the processor, the storage device holding a risk model for calculating a health-related risk value, dictionary information for associating health-related information with at least one of a plurality of aggregation categories for aggregating the health-related information (e.g., at least one of disease name definition information 500, drug definition information 600, surgery definition information 700, hospitalization definition information 800, and test result definition information 900), and explanatory variable definition information for associating one or more combinations of the aggregation categories with values of explanatory variables of the risk model (e.g., explanatory variable definition information 1000), includes a first step (e.g., at least one of steps 1604, 1704, 1804, 1904, and 2004) in which the processor extracts the health-related information from the input information; a second step (e.g., at least one of steps 1605, 1705, 1805, 1905, and 2005) in which the processor identifies one or more aggregation categories corresponding to the extracted health-related information based on the dictionary information; a third step (e.g., steps 2102 to 2105) in which the processor identifies explanatory variable values corresponding to the combination of the identified one or more aggregation categories based on the explanatory variable definition information; and a fourth step (e.g., step 2204) in which the processor calculates the risk value corresponding to the identified explanatory variable value based on the risk model.

This allows the values of notification information items to be appropriately aggregated, and risk assessment based on notification information containing a variety of expressions can be automated using a risk model.

(2) In the above (1), the health-related information includes at least one of information describing the disease name (e.g., disease name 303), information describing the medication (e.g., medication name 304), information about health-related test values (e.g., BMI 403 to HbA1c 405), information describing the surgery (e.g., surgery 307), and information describing the length of hospitalization (e.g., admission date 305 and discharge date 306), and the dictionary information includes at least one of information that associates multiple descriptions about the disease name with one of the aggregate categories (e.g., disease name definition information 500), information that associates multiple descriptions about the medication with one of the aggregate categories (e.g., medication definition information 600), information that associates the range of the health-related test values with the aggregate categories (e.g., test result definition information 900), information that associates multiple descriptions about the surgery with one of the aggregate categories (e.g., surgery definition information 700), and information that associates the range of the length of the hospitalization with the aggregate categories (e.g., hospitalization definition information 800).

This allows for proper aggregation of health-related information.

(3) In the above (2), the input information is information included in notification information submitted by a person who wishes to subscribe to an insurance product, the health risk value is a value indicating the magnitude of risk of insurance payment for the insurance product, the storage device further holds classification definition information (e.g., classification definition information 1500) for classifying the person based on the risk value, and the information analysis support method further includes a fifth step (e.g., steps 2303 to 2304) in which the processor classifies the person based on the calculated risk value and the classification definition information.

This makes it possible to automate the risk assessment of people seeking insurance coverage.

(4) In the above (3), the dictionary information includes, for each of the correspondences between the health-related information and the aggregation category, information indicating whether the identified aggregation category is a target for the calculation of the risk value based on the risk model (e.g., at least one of evaluation processing category (disease name) 503, evaluation processing category (medication) 603, evaluation processing category (surgery) 703, evaluation processing category (hospitalization) 803, and evaluation processing category (examination) 903), and in the second step, the processor determines whether the identified aggregation category is a target for the calculation of the risk value based on the risk model, and if the identified aggregation category is not a target for the calculation of the risk value based on the risk model, the explanatory variable identified based on the identified aggregation category is excluded from the targets of the fourth step and the fifth step.

This ensures a certain level of accuracy in automated risk assessment.

(5) In (4) above, the dictionary information includes information indicating that the identified aggregation category is not subject to calculation of the risk value based on the risk model when the accuracy of the correspondence between the description of the disease name and the aggregation category is lower than a predetermined standard, when the accuracy of the correspondence between the description of the drug and the aggregation category is lower than a predetermined standard, or when the accuracy of the correspondence between the description of the surgery and the aggregation category is lower than a predetermined standard.

This ensures a certain level of accuracy in automated risk assessment.

(6) In (4) above, the dictionary information includes information indicating that the identified aggregation category corresponding to a range of the health-related test values whose frequency of occurrence is lower than a predetermined standard, or the aggregation category corresponding to a range of the length of hospitalization whose frequency of occurrence is lower than a predetermined standard, is not subject to calculation of the risk value based on the risk model.

This ensures a certain level of accuracy in automated risk assessment.

(7) In the above (3), the dictionary information includes information indicating whether confirmation of the identified aggregation category is requested for each of the associations between the health-related information and the aggregation category (e.g., at least one of evaluation processing category (disease name) 503, evaluation processing category (medication) 603, evaluation processing category (surgery) 703, evaluation processing category (hospitalization) 803, and evaluation processing category (examination) 903), and in the second step, the processor determines whether confirmation of the identified aggregation category is requested, and the information analysis support method further includes a sixth step (e.g., step 2305) in which the processor displays the result of classifying the person and, if confirmation of the identified aggregation category is requested, outputs information indicating that confirmation of the identified aggregation category is requested.

This ensures a certain level of accuracy in automated risk assessment.

(8) In (7) above, the dictionary information includes information indicating that confirmation of the identified aggregation category is requested if the degree of correspondence between the description of the disease name and the aggregation category is lower than a predetermined standard, if the degree of correspondence between the description of the drug and the aggregation category is lower than a predetermined standard, or if the degree of correspondence between the description of the surgery and the aggregation category is lower than a predetermined standard.

This ensures a certain level of accuracy in automated risk assessment.

(9) In (7) above, the dictionary information includes information indicating a request to confirm the identified aggregation category, for the aggregation category corresponding to a range of the health-related test values whose frequency of occurrence is lower than a predetermined standard, or for the aggregation category corresponding to a range of the length of hospitalization whose frequency of occurrence is lower than a predetermined standard.

This ensures a certain level of accuracy in automated risk assessment.

(10) An information analysis support method executed by a computer system (e.g., information analysis support system 101), the computer system having a processor (e.g., CPU 104) and a storage device (e.g., storage medium 106 or database 107) connected to the processor, the storage device storing dictionary information (e.g., at least one of disease name definition information 500, drug definition information 600, surgery definition information 700, hospitalization definition information 800, and test result definition information 900) that associates health-related information with at least one of a plurality of aggregation categories that aggregate the health-related information, and a combination of one or more of the aggregation categories and an insurance payment category. The information analysis support method includes a first step (e.g., at least one of steps 3303 to 3307) in which the processor extracts the health-related information from the information claiming payment of insurance money, a second step (e.g., step 3308) in which the processor identifies one or more aggregate categories that correspond to the extracted health-related information based on the dictionary information, and a third step (e.g., step 3308) in which the processor identifies the payment category that corresponds to the combination of the identified one or more aggregate categories based on the payment category definition information.

This allows the values of billing information items to be appropriately aggregated, automating the determination of payment classifications based on billing information that includes a variety of expressions.

(11) In the above (10), the health-related information includes at least one of information describing the disease name (e.g., disease name 303), information describing the medication (e.g., medication name 304), information about health-related test values (e.g., BMI 403 to HbA1c 405), information describing the surgery (e.g., surgery 307), and information describing the length of hospitalization (e.g., admission date 305 and discharge date 306), and the dictionary information includes at least one of information that associates multiple descriptions about the disease name with one of the aggregation categories (e.g., disease name definition information 500), information that associates multiple descriptions about the medication with one of the aggregation categories (e.g., medication definition information 600), information that associates the range of the health-related test values with the aggregation category (e.g., test result definition information 900), information that associates multiple descriptions about the surgery with one of the aggregation categories (e.g., surgery definition information 700), and information that associates the range of the length of the hospitalization with the aggregation category (e.g., hospitalization definition information 800).

This allows for proper aggregation of health-related information.

(12) In the above (11), the dictionary information includes, for each of the correspondences between the health-related information and the aggregation category, information indicating whether the identified aggregation category is a specific target of the payment category based on the payment category definition information (e.g., at least one of evaluation processing category (disease name) 503, evaluation processing category (medication) 603, evaluation processing category (surgery) 703, evaluation processing category (hospitalization) 803, and evaluation processing category (examination) 903), and in the second step, the processor determines whether the identified aggregation category is a specific target of the payment category based on the payment category definition information, and if the identified aggregation category is not a specific target of the payment category based on the payment category definition information, the identified aggregation category is excluded from the target of the third step.

This ensures a certain level of accuracy in automated payment classification determination.

(13) In the above (11), the dictionary information includes information indicating whether confirmation of the identified aggregation category is requested for each of the associations between the health-related information and the aggregation category (e.g., at least one of evaluation processing category (disease name) 503, evaluation processing category (medication) 603, evaluation processing category (surgery) 703, evaluation processing category (hospitalization) 803, and evaluation processing category (examination) 903), and in the second step, the processor determines whether confirmation of the identified aggregation category is requested, and the information analysis support method further includes a fourth step in which the processor displays the identified payment category and, if confirmation of the identified aggregation category is requested, outputs information indicating that confirmation of the identified aggregation category is requested.

This ensures a certain level of accuracy in automated payment classification determination.

The present invention is not limited to the above-described embodiments, but includes various modifications. For example, the above-described embodiments have been described in detail to provide a better understanding of the present invention, and are not necessarily limited to those having all of the configurations described. It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is also possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

The above configurations, functions, processing units, processing means, etc. may be realized in hardware, in part or in whole, for example by designing them as integrated circuits. The above configurations, functions, etc. may be realized in software, by a processor interpreting and executing a program that realizes each function. Information on the programs, tables, files, etc. that realize each function can be stored in storage devices such as non-volatile semiconductor memory, hard disk drives, and SSDs (Solid State Drives), or in computer-readable non-transitory data storage media such as IC cards, SD cards, and DVDs.

In addition, the control lines and information lines shown are those considered necessary for the explanation, and not all control lines and information lines on the product are necessarily shown. In reality, it can be assumed that almost all components are interconnected.

101 Information analysis support system 102 Input unit 103 Output unit 104 CPU
105 Memory 106 Storage medium 107 Database 108 Communication unit 111 Disease name information extraction unit 112 Drug information extraction unit 113 Test result information extraction unit 114 Surgery information extraction unit 115 Hospitalization information extraction unit 116 Explanatory variable information generation unit 117 Risk calculation unit 118 Risk judgment unit 119 Payment judgment unit 121 Basic information management unit 122 Health information management unit 123 Disease name definition information management unit 124 Drug definition information management unit 125 Surgery definition information management unit 126 Hospitalization definition information management unit 127 Test result definition information management unit 128 Explanatory variable definition information management unit 129 Explanatory variable generation result management unit 130 Risk calculation target information management unit 131 Risk model information management unit 132 Classification definition management unit 133 Contract information management unit 134 Billing information management unit 135 Payment classification definition information management unit 136 Payment judgment result information management unit

Claims (10)

An information analysis support method executed by a computer system, comprising:
The computer system includes a processor and a storage device connected to the processor;
The storage device includes:
A risk model for calculating a risk value for health;
Dictionary information that associates information related to health with at least one of a plurality of aggregation categories that aggregate the information related to health;
retaining explanatory variable definition information that associates one or more combinations of the aggregation categories with values of explanatory variables of the risk model;
The information analysis support method includes:
A first step in which the processor extracts the health-related information from input information;
a second step of the processor identifying one or more of the aggregation categories corresponding to the extracted health-related information based on the dictionary information;
a third step of the processor identifying values of the explanatory variables corresponding to the identified combination of one or more aggregation categories based on the explanatory variable definition information;
and a fourth step of the processor calculating the risk value corresponding to the value of the identified explanatory variable based on the risk model. 2. The information analysis support method according to claim 1,
The health-related information includes at least one of information describing a disease name, information describing a medication, information describing a test result related to health, information describing an operation, and information describing a length of hospitalization;
An information analysis support method characterized in that the dictionary information includes at least one of information that corresponds multiple descriptions of the disease name to one of the aggregation categories, information that corresponds multiple descriptions of the medication to one of the aggregation categories, information that corresponds a range of test values related to the health to the aggregation categories, information that corresponds multiple descriptions of the surgery to one of the aggregation categories, and information that corresponds a range of the length of the hospital stay to the aggregation categories. 3. The information analysis support method according to claim 2,
The input information is information included in notification information submitted by a person who wishes to subscribe to an insurance product,
the health risk value is a value indicating the magnitude of risk of insurance payment for the insurance product,
the storage device further stores classification definition information for classifying the person based on the risk value;
The information analysis support method, further comprising a fifth step in which the processor classifies the person based on the calculated risk value and the classification definition information. 4. The information analysis support method according to claim 3,
the dictionary information includes, for each of the associations between the health-related information and the aggregation categories, information indicating whether the identified aggregation category is a target for calculation of the risk value based on the risk model;
In the second step, the processor determines whether the identified aggregation category is a target for calculating the risk value based on the risk model;
An information analysis support method, characterized in that if the identified aggregation category is not subject to calculation of the risk value based on the risk model, the explanatory variable identified based on the identified aggregation category is excluded from the subjects of the fourth step and the fifth step. 5. The information analysis support method according to claim 4,
An information analysis support method characterized in that the dictionary information includes information indicating that the identified aggregation category is not subject to calculation of the risk value based on the risk model when the accuracy of the correspondence between the description of the disease name and the aggregation category is lower than a predetermined standard, when the accuracy of the correspondence between the description of the medication and the aggregation category is lower than a predetermined standard, or when the accuracy of the correspondence between the description of the surgery and the aggregation category is lower than a predetermined standard. 5. The information analysis support method according to claim 4,
The information analysis support method is characterized in that the dictionary information includes information indicating that the identified aggregation category corresponding to a range of the health-related test values whose frequency of occurrence is lower than a predetermined standard, or the aggregation category corresponding to a range of the length of hospitalization whose frequency of occurrence is lower than a predetermined standard, is not subject to calculation of the risk value based on the risk model. 4. The information analysis support method according to claim 3,
The dictionary information includes information indicating whether to request confirmation of the identified aggregation category for each of the associations between the health-related information and the aggregation category,
In the second step, the processor determines whether confirmation of the identified aggregation category is required;
The information analysis support method is characterized in that it further includes a sixth step in which the processor displays the results of classifying the person, and, if confirmation of the identified aggregation category is requested, outputs information indicating that confirmation of the identified aggregation category is requested. The information analysis support method according to claim 7,
An information analysis support method characterized in that the dictionary information includes information indicating that confirmation of the identified aggregation category is requested when the accuracy of the correspondence between the description of the disease name and the aggregation category is lower than a predetermined standard, when the accuracy of the correspondence between the description of the medication and the aggregation category is lower than a predetermined standard, or when the accuracy of the correspondence between the description of the surgery and the aggregation category is lower than a predetermined standard. The information analysis support method according to claim 7,
An information analysis support method characterized in that the dictionary information includes information indicating a request for confirmation of the identified aggregation category, for the aggregation category corresponding to a range of the health-related test values whose frequency of occurrence is lower than a predetermined standard, or for the aggregation category corresponding to a range of the length of hospitalization whose frequency of occurrence is lower than a predetermined standard. An information analysis support system,
A processor and a storage device connected to the processor,
The storage device includes:
A risk model for calculating a risk value for health;
Dictionary information that associates information related to health with at least one of a plurality of aggregation categories that aggregate the information related to health;
retaining explanatory variable definition information that associates one or more combinations of the aggregation categories with values of explanatory variables of the risk model;
The processor,
Extracting the health-related information from the input information;
Identifying one or more of the aggregation categories corresponding to the extracted health-related information based on the dictionary information;
Identifying values of the explanatory variables corresponding to the combination of the one or more aggregation categories identified based on the explanatory variable definition information;
An information analysis support system, comprising: a risk model that calculates the risk value corresponding to the value of the specified explanatory variable.

Images