Medical checkup recommendation system, medical checkup recommendation method, and program

Abstract

To provide a medical checkup recommendation system, a medical checkup recommendation method, and a program capable of improving a medical checkup reception rate by medical checkup recommendation.SOLUTION: A set of first data indicating presence or absence of medical checkup of each subject in a plurality of years before a reference year and second data indicating presence or absence of medical checkup of each subject in the reference year is defined as one data set. A medical checkup recommendation system includes a prediction unit that predicts reception probability that each subject receives medical checkup based on a relation between the first data and the second data in the plurality of data sets of which the reference year is different from each other.SELECTED DRAWING: Figure 8

Description

[Technical field]

[0001] The present invention relates to a health checkup recommendation system, a health checkup recommendation method, and a program. [Background technology]

[0002] It is considered important for the nation's citizens to undergo health checkups in order to maintain their health and prevent diseases, and insurers and businesses are required by law to implement specific health checkups and medical examinations. However, for those who do not undergo health checkups, or those who undergo health checkups but are not interested in the results, there is a risk that they will delay visiting a medical institution and miss an opportunity for appropriate treatment. In order to increase the rate of people undergoing health checkups and to lead to them visiting a medical institution, recommendations for undergoing health checkups are being made.

[0003] For example, Patent Document 1 listed below discloses a technology that identifies medical examination institutions where a subject is recommended to undergo a medical examination based on the subject's medical examination history, and sends a message to the subject's mobile device recommending that the subject undergo a medical examination at an examination institution located within a specified range of the subject. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2018-124836 A Summary of the Invention [Problem to be solved by the invention]

[0005] The technology described in the above Patent Document 1 encourages people to take a health check without distinguishing between people who have taken a health check and people who have not. However, among people who have not taken a health check, there are people who do not take a health check even if they are encouraged to take a health check. For this reason, even if people are encouraged to take a health check without distinguishing between people who have taken a health check and people who have not, it is not possible to increase the rate of people taking a health check.

[0006] In view of the above-mentioned problems, an object of the present invention is to provide a health check recommendation system, a health check recommendation method, and a program that can improve the health check-up attendance rate through health check-up recommendation. [Means for solving the problem]

[0007] In order to solve the above-mentioned problems, a health checkup recommendation system according to one embodiment of the present invention is a health checkup recommendation system that includes a prediction unit that treats a set of first data indicating whether or not each subject has received a health checkup for multiple years prior to a base year and second data indicating whether or not each subject has received a health checkup in the base year as one dataset, and predicts the probability that each of the subjects will receive a health checkup based on the relationship between the first data and the second data in the multiple datasets with different base years.

[0008] A health checkup recommendation method according to one embodiment of the present invention is a computer-executed health checkup recommendation method including a prediction process for predicting the probability of each of the subjects undergoing a health checkup based on the relationship between the first data and the second data in multiple datasets having different base years, the multiple datasets being a set of first data indicating whether or not each of the subjects has undergone a health checkup for multiple years prior to a base year and second data indicating whether or not each of the subjects has undergone a health checkup in the base year.

[0009] A program according to one embodiment of the present invention is a program for causing a computer to function as a prediction means for predicting the probability of each of the subjects undergoing a health checkup based on the relationship between the first data and the second data in multiple datasets having different base years, the multiple datasets being a set of first data indicating whether or not each of the subjects has undergone a health checkup in multiple years prior to a base year and second data indicating whether or not each of the subjects has undergone a health checkup in the base year. Effect of the Invention

[0010] According to the present invention, it is possible to improve the rate of people receiving health checkups by encouraging them to receive health checkups. [Brief description of the drawings]

[0011] [Figure 1] 1 is a block diagram showing an example of the configuration of a health check recommendation system according to an embodiment of the present invention. [Diagram 2] 1 is a diagram showing an overview of processing performed by a health checkup recommendation system according to an embodiment of the present invention. [Diagram 3] FIG. 2 is a block diagram showing an example of a functional configuration of a learning server according to the present embodiment. [Figure 4] FIG. 11 is a diagram illustrating an example of teacher data generation according to the present embodiment. [Diagram 5] FIG. 11 is a diagram illustrating an example of teacher data generation according to the present embodiment. [Figure 6] FIG. 11 is a diagram illustrating an example of teacher data generation according to the present embodiment. [Figure 7] FIG. 11 is a diagram illustrating an example of teacher data generation according to the present embodiment. [Figure 8] 2 is a block diagram showing an example of a functional configuration of a health check recommendation server according to the present embodiment. FIG. [Figure 9] 10 is a flowchart showing an example of a processing flow in a learning server according to the present embodiment. [Figure 10] 10 is a flowchart showing an example of a process flow in a health checkup recommendation server according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0012] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[0013] <1. Structure of the health checkup recommendation system> The configuration of a health check recommendation system according to this embodiment will be described with reference to Figures 1 and 2. Figure 1 is a block diagram showing an example of the configuration of a health check recommendation system according to this embodiment. The health checkup recommendation system shown in Figure 1 is a system that predicts the probability that a subject will undergo a health checkup (hereinafter also referred to as "probability of undergoing a health checkup"), and generates and outputs a list (hereinafter also referred to as "recommender list") indicating those who are to be recommended to undergo a health checkup (hereinafter also referred to as "recommenders") based on the probability of undergoing a health checkup. For example, subjects with a high probability of undergoing a health checkup are preferentially selected as viewers indicated by the recommender list. For this reason, a person who recommends recommenders to undergo a health checkup (hereinafter also referred to as "recommender") can improve the rate of people undergoing a health checkup by recommending the viewers indicated by the recommender list to undergo a health checkup. The target person is, for example, a person who is enrolled in health insurance (insured person). In this case, the recommender is an insured person selected from among the insured persons based on the probability of receiving medical treatment. The recommender is, for example, a local government or a health insurance association. A health checkup is a medical examination (health check-up) and is classified into general health checkups (regular health check-ups) and specific health checkups.

[0014] As shown in FIG. 1, the health checkup recommendation system 1 includes a user terminal 10, a learning server 20, and a health checkup recommendation server 30. The network NW may be configured to transmit and receive information using, for example, a Local Area Network (LAN), a Wide Area Network (WAN), a telephone network (mobile phone network, landline telephone network, etc.), a regional Internet Protocol (IP) network, the Internet, etc.

[0015] (1) User terminal 10 The user terminal 10 is a terminal owned by a user. The user is, for example, a person in charge (recommendation implementer) of a local government, a health insurance association, or the like. The user terminal 10 is, for example, a mobile terminal such as a smartphone or a tablet terminal, or a PC (Personal Computer). The user terminal 10 is communicably connected to the learning server 20 and the health check recommendation server 30 via a network NW.

[0016] (2) Learning Server 20 The learning server 20 is a server that executes a learning process. In the learning process, learning is performed to predict the probability of medical examination, and a model that predicts the probability of medical examination (hereinafter also referred to as a "prediction model") is generated. The learning server 20 is composed of one or more servers (e.g., cloud servers). The learning server 20 is communicatively connected to the user terminal 10 and the health check recommendation server 30 via the network NW.

[0017] (3) Health checkup recommendation server 30 The health checkup recommendation server 30 is a server that executes processing for recommending health checkups. The health checkup recommendation server 30 is composed of one or more servers (e.g., cloud servers). The health checkup recommendation server 30 is connected to the user terminal 10 and the learning server 20 via the network NW so as to be able to communicate with them.

[0018] Here, an overview of the processing by the health checkup recommendation system 1 will be described with reference to Fig. 2. Fig. 2 is a diagram showing an overview of the processing by the health checkup recommendation system 1 according to this embodiment.

[0019] As shown in FIG. 2, subject information indicating information about a subject is input as input data to the health check recommendation system 1. The subject information is information indicating, for example, the subject's medical examination history, gender, age, residential area, etc. The subject's medical examination history is information indicating whether or not the subject has undergone a health checkup conducted in the past several years (whether or not the subject has undergone a health checkup). Information indicating the subject's medical examination history is also referred to as "medical examination history information" below.

[0020] When subject information is input, the health checkup recommendation system 1 first performs segment classification (step S1). Specifically, the health checkup recommendation system 1 classifies each subject into a segment according to the characteristics of the subject indicated in the input subject information. For example, the health checkup recommendation system 1 classifies subjects into segments according to gender, age, residential area, etc. Note that segments that combine multiple characteristics (for example, segments according to gender and age) may be prepared.

[0021] After segment classification, the health checkup recommendation system 1 predicts the probability of undergoing a medical checkup using a prediction model (step S2). At this time, the health checkup recommendation system 1 predicts the probability of undergoing a medical checkup using a prediction model for each segment into which the subject has been classified, and summarizes the prediction results for each segment.

[0022] After predicting the probability of receiving a medical checkup, the health checkup recommendation system 1 narrows down the candidates for recommendation based on the prediction result (step S3). At this time, the health checkup recommendation system 1 narrows down the candidates for recommendation based on the segment into which the subjects are classified. For example, the health checkup recommendation system 1 preferentially extracts as candidates those with a high predicted probability of receiving a medical checkup from among the subjects for each segment. A subject with a high probability of receiving a medical checkup is, for example, a subject whose probability of receiving a medical checkup is equal to or higher than a predetermined reference value.

[0023] After narrowing down the recommenders, the health checkup recommendation system 1 outputs a list of recommenders (step S4). At this time, the health checkup recommendation system 1 generates and outputs a list of recommenders for each segment into which the subjects are classified. For example, the health checkup recommendation system 1 generates and outputs a list indicating the recommenders extracted for each segment and their probability of undergoing checkup as a list of recommenders, as shown in Figure 2.

[0024] <2. Functional configuration of the learning server> The configuration of the health checkup recommendation system 1 according to this embodiment has been described above. Next, the functional configuration of the learning server 20 according to this embodiment will be described with reference to Figures 3 to 7. Figure 3 is a block diagram showing an example of the functional configuration of the learning server 20 according to this embodiment. As shown in FIG. 3, the learning server 20 includes a communication unit 210, a storage unit 220, and a control unit 230.

[0025] (1) Communications unit 210 The communication unit 210 has a function of transmitting and receiving various information. The communication unit 210 communicates with the user terminal 10 via the network NW and receives medical examination history information. The communication unit 210 also communicates with the health check recommendation server 30 via the network NW and transmits a prediction model.

[0026] (2) Storage section 220 The storage unit 220 has a function of storing various information. The storage unit 220 is configured by a storage medium provided as hardware in the learning server 20, such as a hard disk drive (HDD), a solid state drive (SSD), a flash memory, an electrically erasable programmable read only memory (EEPROM), a random access read / write memory (RAM), a read only memory (ROM), or any combination of these storage media. As shown in FIG. 3, the storage unit 220 includes a medical examination history information storage unit 221, a teacher data storage unit 222, and a prediction model storage unit 223.

[0027] (2-1) Medical History Information Storage Unit 221 The medical examination history information storage unit 221 has a function of storing medical examination history information. The medical examination history information is information that the communication unit 210 receives from the user terminal 10.

[0028] (2-2) Teacher data storage unit 222 The teacher data storage unit 222 has a function of storing teacher data. The teacher data is data generated by the data processing unit 232 described later. Note that the teacher data is not limited to data generated by the data processing unit 232, and may be data generated by the user terminal 10 and received by the communication unit 210 from the user terminal 10.

[0029] (2-3) Prediction model storage unit 223 The prediction model storage unit 223 has a function of storing a prediction model. The prediction model is a model generated by the learning unit 233, which will be described later.

[0030] (3) Control unit 230 The control unit 230 has a function of controlling the overall operation of the learning server 20. The control unit 230 is realized, for example, by causing a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) that the learning server 20 has as hardware to execute a program. As shown in FIG. 3, the control unit 230 includes a data acquisition unit 231, a data processing unit 232, a learning unit 233, and an output processing unit 234.

[0031] (3-1) Data acquisition unit 231 The data acquisition unit 231 has a function of acquiring various data. For example, the data acquisition unit 231 acquires medical examination history information received by the communication unit 210 from the user terminal 10, and stores the medical examination history information in the medical examination history information storage unit 221.

[0032] (3-2) Data processing unit 232 The data processing unit 232 has a function of processing various data. For example, the data processing unit 232 performs a process of generating teacher data from the medical examination history information acquired by the data acquisition unit 231. After the generation, the data processing unit 232 stores the generated teacher data in the teacher data storage unit 222. The data processing unit 232 may access the medical examination history information storage unit 221 by itself to acquire the medical examination history information.

[0033] First, the data processing unit 232 sets a reference year in the medical examination history information. The reference year is a year that is a reference for generating teacher data. Next, the data processing unit 232 extracts, as explanatory variable data, first data indicating whether or not each subject has had a medical checkup in the past several years before the set base year. Also, the data processing unit 232 extracts, as correct answer data, second data indicating whether or not each subject has had a medical checkup in the base year. Next, the data processing unit 232 generates one data set consisting of the extracted first data and second data as one teacher data. When generating multiple teacher data, the data processing unit 232 first generates teacher data using one year ago as the base year, shifts the base year to two years ago, three years ago, and so on, and generates teacher data for each of the set base years. In this way, the data processing unit 232 generates multiple teacher data.

[0034] Here, the teacher data generated by the data processing unit 232 will be described with reference to Fig. 4 to Fig. 7. Fig. 4 to Fig. 7 are diagrams showing an example of teacher data generation according to this embodiment. In the examples shown in Figures 4 to 6, for ease of explanation, an example is shown in which explanatory variable data for the past three years from the base year is extracted based on the medical examination history information of subjects with IDs 1 to 6. In the example shown in Figure 7, for ease of explanation, an example is shown in which explanatory variable data for the past three years from the base year is extracted based on the medical examination history information of subjects with IDs 7 to 8. In the examples shown in Figures 4 to 7, the presence of a medical examination is indicated by "◯" and the absence of a medical examination is indicated by "X".

[0035] Fig. 4 shows an example of generating teacher data 40 when the base year is set to one year ago. As shown in Fig. 4, teacher data 40 is generated by combining explanatory variable data 41 indicating whether or not a person underwent a health checkup two to four years ago (the past three years) and correct answer data 42 indicating whether or not a person underwent a health checkup one year ago.

[0036] Fig. 5 shows an example of teacher data 40 generated when the base year is set to two years ago. As shown in Fig. 5, teacher data 40 is generated by combining explanatory variable data 41 indicating whether or not a person underwent a health checkup three to five years ago (the past three years) and correct answer data 42 indicating whether or not a person underwent a health checkup two years ago.

[0037] Fig. 6 shows an example of teacher data 40 generated when the base year is set to three years ago. As shown in Fig. 6, teacher data 40 is generated by combining explanatory variable data 41 indicating whether or not a person underwent a health checkup four to six years ago (the past three years) and correct answer data 42 indicating whether or not a person underwent a health checkup three years ago.

[0038] Fig. 7 shows an example of teacher data 40 generated when the base year is set to one year ago. As shown in Fig. 7, teacher data 40 is generated by combining explanatory variable data 41 indicating whether or not a health check was taken two to four years ago (the past three years) and correct answer data 42 indicating whether or not a health check was taken one year ago.

[0039] In generating training data, there may be cases where all correct answer data is generated as no medical checkup (or yes), as in the example shown in Figure 7. If only whether or not a health checkup was attended in the past year is used as the correct answer data, learning based on training data such as that shown in Figure 7 may have a significant impact on the prediction model's prediction of the probability of attending a health checkup, and may significantly reduce its accuracy. In contrast, in the health checkup recommendation system 1 according to this embodiment, multiple pieces of teacher data are generated by shifting the base year every year. This eliminates the influence of "human behavior caused by many factors intertwined in a complex manner" that occurred only in the most recent year, and minimizes the influence of teacher data such as that shown in Fig. 7 on the prediction of the probability of attending a health checkup by the prediction model, thereby ensuring the accuracy of the prediction. In particular, whether or not to attend a health check is determined by a "complex interplay of many factors" rather than "physical condition," and by treating "whether or not to attend a health check" as "data with explanatory variables," the accuracy of predicting the probability of attending a health check can be further improved.

[0040] (3-3) Learning Section 233 The learning unit 233 has a function of generating a prediction model (trained model) by learning. The learning unit 233 generates a prediction model that has learned the relationship between explanatory variable data (first data) and correct answer data (second data) based on multiple teacher data generated by the data processing unit 232. The prediction model that has learned the relationship receives medical examination history information of the subject as an input, and outputs the probability that the subject will receive a medical examination. After generation, the learning unit 233 stores the generated prediction model in the prediction model storage unit 223.

[0041] (3-4) Output processing unit 234 The output processing unit 234 has a function of controlling the output of various information. For example, the output processing unit 234 transmits the prediction model generated by the learning unit 233 to the health checkup recommendation server 30 via the communication unit 210.

[0042] <3. Functional configuration of the health check recommendation server> The functional configuration of the learning server 20 according to this embodiment has been described above. Next, the functional configuration of the health checkup recommendation server 30 according to this embodiment will be described with reference to Fig. 8. Fig. 8 is a block diagram showing an example of the functional configuration of the health checkup recommendation server 30 according to this embodiment. As shown in FIG. 8, the health checkup recommendation server 30 includes a communication unit 310, a storage unit 320, and a control unit 330.

[0043] (1) Communications Unit 310 The communication unit 310 has a function of transmitting and receiving various information. The communication unit 310 communicates with the user terminal 10 via the network NW to receive target person information and transmit a recommender list. The communication unit 310 also communicates with the learning server 20 via the network NW to receive a prediction model.

[0044] (2) Storage section 320 The storage unit 320 has a function of storing various information. The storage unit 320 is configured by a storage medium provided as hardware in the health checkup recommendation server 30, such as an HDD, SSD, flash memory, EEPROM, RAM, ROM, or any combination of these storage media. As shown in FIG. 8, the storage unit 320 includes a prediction model storage unit 321 and a subject information storage unit 322.

[0045] (2-1) Prediction model storage unit 321 The prediction model storage unit 321 has a function of storing a prediction model. The prediction model is generated by the learning server 20 and received by the communication unit 310 from the learning server 20.

[0046] (2-2) Target person information storage unit 322 The subject information storage unit 322 has a function of storing subject information. The subject information is information that the communication unit 310 receives from the user terminal 10.

[0047] (3) Control unit 330 The control unit 330 has a function of controlling the overall operation of the health checkup recommendation server 30. The control unit 330 is realized, for example, by causing a CPU or GPU provided as hardware in the health checkup recommendation server 30 to execute a program. As shown in FIG. 8, the control unit 330 includes a subject information acquisition unit 331, a segment classification unit 332, a prediction unit 333, a narrowing down unit 334, a recommendation processing unit 335, and an output processing unit 336.

[0048] (3-1) Target person information acquisition unit 331 The subject information acquisition unit 331 has a function of acquiring subject information. The subject information acquisition unit 331 acquires the subject information received by the communication unit 310 from the user terminal 10, and stores the subject information in the subject information storage unit 322.

[0049] (3-2) Segment classification unit 332 The segment classification unit 332 has a function of classifying subjects into segments. For example, the segment classification unit 332 classifies each subject into a segment according to the characteristics of the subject, based on the characteristics of the subject indicated by the subject information acquired by the subject information acquisition unit 331. Specifically, the segment classification unit 332 classifies the subjects into a segment according to gender, a segment according to age, a segment according to residential area, etc. In addition, the segment classification unit 332 may classify the subjects into a segment that combines multiple characteristics, such as a segment according to gender and age, a segment according to gender and residential area, a segment according to age and residential area, or a segment according to gender, age and residential area.

[0050] (3-3) Prediction Section 333 The prediction unit 333 has a function of predicting the probability of a subject receiving a health checkup. The prediction unit 333 predicts the probability of each subject receiving a health checkup based on the relationship between the first data and the second data in a plurality of data sets having different base years. For example, the prediction unit 333 predicts the probability of receiving a health checkup using a prediction model that has learned the relationship between the first data and the second data, using a data set in which the first data is explanatory variable data and the second data is correct answer data as teacher data. The prediction model is generated, for example, by learning in the learning server 20, and stored in the prediction model storage unit 321 of the health checkup recommendation server 30.

[0051] When predicting the probability of consultation using the prediction model, the prediction unit 333 acquires necessary medical examination history information from the subject information acquired by the subject information acquisition unit 331 and inputs it to the prediction model. At this time, the prediction unit 333 acquires the medical examination history information of the subject for each segment classified by the segment classification unit 332 and inputs it to the prediction model. The prediction unit 333 obtains the visit probability output from the prediction model by inputting the medical examination history information as a prediction result. At this time, the prediction unit 333 manages the prediction results collectively for each segment. In this way, the prediction unit 333 predicts the visit probability for each segment.

[0052] (3-4) Filtering section 334 The narrowing down unit 334 has a function of narrowing down recommenders. The narrowing down unit 334 narrows down recommenders by extracting, from among the targets, targets who have a high probability of receiving treatment as targets for recommendation. At this time, the narrowing down unit 334 narrows down the recommenders for each segment based on the prediction result by the prediction unit 333 of the probability of receiving treatment for each segment. Specifically, the selection unit 334 preferentially extracts subjects with a high predicted probability of receiving medical treatment as recommended subjects from among the subjects for each segment. As an example, the selection unit 334 extracts subjects with a probability of receiving medical treatment of 50% or more (a predetermined reference value) as subjects with a high probability of receiving medical treatment. On the other hand, the selection unit 334 does not extract subjects with a probability of receiving medical treatment of less than 50% as subjects with a high probability of receiving medical treatment. In other words, the selection unit 334 excludes subjects with a low probability of receiving medical treatment from the recommendation targets. The specific numerical value of the predetermined reference value is not limited to 50%.

[0053] (3-5) Recommendation Processing Section 335 The recommendation processing unit 335 has a function of generating a recommendation list. The recommendation processing unit 335 generates a list of encouragers based on the consultation probability predicted by the prediction unit 333. Specifically, the recommendation processing unit 335 generates a list of encouragers indicating the consultation probability for each encourager extracted by the selection unit 334. At this time, the recommendation processing unit 335 generates a list of encouragers for each segment into which each encourager is classified.

[0054] (3-6) Output processing unit 336 The output processing unit 336 has a function of controlling the output of various information. For example, the output processing unit 336 transmits the recommender list generated by the recommendation processing unit 335 to the user terminal 10 via the communication unit 310. This allows the recommendation provider to confirm those recommenders who are highly likely to undergo a health check by displaying the recommender list on the user terminal 10. The recommendation provider can increase the rate of people taking the health check by giving priority to the recommendation tasks to those recommended by those confirmed on the recommender list.

[0055] <4. Processing flow> The functional configuration of the health checkup recommendation server 30 according to this embodiment has been described above. Next, the flow of processing according to this embodiment will be described with reference to Figs.

[0056] (1) Processing flow in the learning server 20 The flow of processing in the learning server 20 will be described with reference to Fig. 9. Fig. 9 is a flowchart showing an example of the flow of processing in the learning server 20 according to this embodiment.

[0057] As shown in FIG. 9, first, the data acquisition unit 231 of the learning server 20 acquires medical examination history information (step S101).

[0058] Next, the data processing unit 232 of the learning server 20 generates teacher data based on the medical examination history information acquired by the data acquisition unit 231 (step S102). Specifically, the data processing unit 232 sets a number of different base years and generates teacher data for each of the set base years, thereby generating a number of teacher data.

[0059] Next, the learning unit 233 of the learning server 20 performs learning using the multiple pieces of teacher data generated by the data processing unit 232 (step S103). Through this learning, the learning unit 233 generates a prediction model.

[0060] After generating the prediction model, the learning unit 233 saves (stores) the generated prediction model in the prediction model storage unit 223 (step S104).

[0061] After saving the prediction model, the output processing unit 234 of the learning server 20 transmits (outputs) the prediction model to the health checkup recommendation server 30 via the communication unit 210 (step S105).

[0062] (2) Processing flow in the health checkup recommendation server 30 The flow of processing in the health checkup recommendation server 30 will be described with reference to Fig. 10. Fig. 10 is a flow chart showing an example of the flow of processing in the health checkup recommendation server 30 according to this embodiment.

[0063] 10, first, the subject information acquisition unit 331 of the health checkup recommendation server 30 acquires subject information (step S201). The subject information is, for example, information indicating a subject for whom the probability of receiving health checkup is to be predicted, transmitted from the user terminal 10 and received by the communication unit 310.

[0064] Next, the segment classification unit 332 of the health checkup recommendation server 30 classifies the subjects into segments according to their characteristics, based on the subject information acquired by the subject information acquisition unit 331 (step S202).

[0065] Next, the prediction unit 333 of the health checkup recommendation server 30 predicts the subject's probability of receiving medical checkup for each segment classified by the segment classification unit 332 (step S203). Specifically, the prediction unit 333 extracts the subject's medical checkup history information for each segment from the subject information acquired by the subject information acquisition unit 331, inputs the medical checkup history information to the prediction model stored in the prediction model storage unit 321, and acquires the output probability of receiving medical checkup as the prediction result.

[0066] After the prediction, the selection unit 334 of the health checkup recommendation server 30 narrows down the recommenders for each segment classified by the segment classification unit 332 (step S204). Specifically, the selection unit 334 narrows down the recommenders by preferentially extracting subjects with a high predicted probability of receiving medical examination from among the subjects for each segment.

[0067] After narrowing down, the recommendation processing unit 335 of the health checkup recommendation server 30 generates a list of recommenders for each segment classified by the segment classification unit 332 (step S205). Specifically, the recommendation processing unit 335 generates a list of recommenders indicating the probability of each recommender receiving medical examination for the recommenders extracted for each segment by the narrowing down unit 334.

[0068] After generation, the output processing unit 336 of the health checkup recommendation server 30 outputs the recommender list for each segment classified by the segment classification unit 332 (step S206). Specifically, the output processing unit 336 transmits the recommender list generated by the recommendation processing unit 335 to the user terminal 10 via the communication unit 310.

[0069] As described above, the health checkup recommendation system 1 of this embodiment includes a prediction unit 333 that predicts the probability of each subject undergoing a health checkup based on the relationship between the first data and the second data in multiple datasets with different base years, using a set of first data indicating whether or not each subject has undergone a health checkup in multiple years prior to a base year and second data indicating whether or not each subject has undergone a health checkup in the base year as one dataset.

[0070] With this configuration, the recommendation implementer can exclude from the recommendation target those subjects with a low predicted probability of receiving a medical checkup as those who will not receive a medical checkup even if they are recommended to receive a medical checkup. Also, the recommendation implementer can increase the medical checkup attendance rate by recommending to receive a medical checkup only those subjects with a high predicted probability of receiving a medical checkup. Therefore, the health checkup recommendation system 1 according to this embodiment makes it possible to improve the health checkup attendance rate by encouraging people to take the health checkup. In addition, if the recommendation to undergo health checkups is made by mail, the cost of mailing the recommendation letter can be reduced by not mailing the recommendation letter to those who, despite being recommended to undergo a health checkup, still do not undergo the checkup.

[0071] <5. Modifications> The above describes the embodiments of the present invention. Next, modified examples of the above-described embodiments will be described. Each modified example described below may be applied to the embodiment alone or in combination with the embodiment. Each modified example may be applied in place of the configuration described in the embodiment, or may be applied in addition to the configuration described in the embodiment.

[0072] In the above-described embodiment, an example has been described in which the health checkup recommendation system 1 includes the user terminal 10, the learning server 20, and the health checkup recommendation server 30, but the present invention is not limited to this example. The health checkup recommendation system 1 may be realized by one device, by multiple devices, or by a combination of multiple devices of different types. For example, the health checkup recommendation system 1 may be realized only by a user terminal 10 having the functions of the learning server 20 and the health checkup recommendation server 30. The health checkup recommendation system 1 may also be realized by a user terminal 10 and one server having the functions of the learning server 20 and the health checkup recommendation server 30. The health checkup recommendation system 1 may also be realized by a user terminal 10 having the functions of either the learning server 20 or the health checkup recommendation server 30, and one or more servers having functions of the learning server 20 or the health checkup recommendation server 30 that the user terminal 10 does not have.

[0073] In the above-described embodiment, an example in which the explanatory variable data (first data) is data indicating only whether or not a medical checkup has been taken has been described, but the present invention is not limited to such an example. For example, the explanatory variable data may include data on receipts for each subject in addition to whether or not a medical checkup has been taken. The data on receipts is, for example, information indicating the number of receipts and the number of hospital visits per year of the subject. For example, it can be said that a subject with a larger number of receipts and the number of hospital visits per year is more concerned about his / her own health condition, and a subject with a smaller number of receipts and the number of hospital visits per year is less concerned about his / her own health condition. For this reason, the medical checkup recommendation system 1 learns that a large number of receipts and the number of hospital visits per year is one factor that calculates a high probability of taking the medical checkup, and that a small number of receipts and the number of hospital visits per year is one factor that calculates a low probability of taking the medical checkup. This enables the health checkup recommendation system 1 to further improve the accuracy of predicting the probability of receiving a health checkup.

[0074] The above describes the modified embodiment of the present invention. The medical checkup recommendation system 1, the learning server 20, and the medical checkup recommendation server 30 in the above-mentioned embodiment may be realized in part or in whole by a computer. The computer here includes a quantum computer. In that case, the function may be realized by recording a program for realizing the function in a computer-readable recording medium, reading the program recorded in the recording medium into a computer system, and executing the program. The term "computer system" here includes hardware such as an OS and peripheral devices. The term "computer-readable recording medium" refers to portable media such as flexible disks, optical magnetic disks, ROMs, and CD-ROMs, and storage devices such as hard disks built into a computer system. The term "computer-readable recording medium" may also include a medium that dynamically holds a program for a short period of time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, and a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that is a server or client in that case. Furthermore, the above program may be for realizing part of the functions described above, or may be capable of realizing the functions described above in combination with a program already recorded in a computer system, or may be realized using a programmable logic device such as an FPGA (Field Programmable Gate Array).

[0075] Although the embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to the above, and various design changes, etc. are possible within the scope that does not deviate from the gist of the present invention. [Explanation of symbols]

[0076] 1...health checkup recommendation system, 10...user terminal, 20...learning server, 30...health checkup recommendation server, 210...communication unit, 220...storage unit, 221...medical examination history information storage unit, 222...teacher data storage unit, 223...prediction model storage unit, 230...control unit, 231...data acquisition unit, 232...data processing unit, 233...learning unit, 234...output processing unit, 310...communication unit, 320...storage unit, 321...prediction model storage unit, 322...subject information storage unit, 330...control unit, 331...subject information acquisition unit, 332...segment classification unit, 333...prediction unit, 334...narrowing unit, 335...recommendation processing unit, 336...output processing unit, NW...network

Claims

1. a prediction unit which predicts the probability of each of the subjects receiving a health checkup based on a relationship between the first data and the second data in the multiple data sets having different base years, the multiple data sets being a set of first data indicating whether or not each of the subjects has received a health checkup in multiple years prior to a base year and second data indicating whether or not each of the subjects has received a health checkup in the base year; A health checkup recommendation system that includes the following:

2. the prediction unit predicts the probability of consultation using a prediction model that has learned a relationship between the first data and the second data, using the data set, in which the first data is explanatory variable data and the second data is correct answer data, as teacher data. The health checkup recommendation system according to claim 1.

3. The first data includes data regarding prescriptions for each subject, The health checkup recommendation system according to claim 1.

4. A segment classification unit that classifies each subject into a segment according to the characteristics of the subject based on the characteristics of the subject; Further equipped with The prediction unit predicts the probability of consultation for each of the segments. The health checkup recommendation system according to claim 1.

5. a recommendation processing unit that generates a list of subjects who are to be recommended to undergo a health checkup based on the predicted probability of undergoing a health checkup; The health checkup recommendation system according to claim 1 , further comprising:

6. a screening unit that extracts, from among the subjects, subjects who have a high probability of receiving medical treatment as subjects to be recommended; Further equipped with the recommendation processing unit generates the list of the recommendation targets extracted by the selection unit. The health checkup recommendation system according to claim 5.

7. The recommendation processing unit generates the list for each segment into which each target person is classified. The health checkup recommendation system according to claim 5.

8. a prediction process for predicting the probability of each of the subjects receiving a health checkup based on a relationship between the first data and the second data in the multiple data sets having different base years, the multiple data sets being a set of first data indicating whether or not each of the subjects has received a health checkup in multiple years prior to a base year and second data indicating whether or not each of the subjects has received a health checkup in the base year; A health check-up recommendation method executed by a computer, comprising:

9. Computer, a prediction means for predicting the probability of each of the subjects receiving a health checkup based on a relationship between the first data and the second data in the multiple data sets having different base years, the multiple data sets being a set of first data indicating whether or not each of the subjects has received a health checkup in multiple years prior to a base year and second data indicating whether or not each of the subjects has received a health checkup in the base year; A program to function as a

Images