Information processing device, support method, and support program

Abstract

Simplifying the process of claiming insurance benefits. [Solution] The information processing device includes an acquisition means for acquiring event information indicating target events that may constitute grounds for insurance payment, and an extraction unit that uses a machine learning-trained extraction model to take a pair of documents and data as input and output the portion of the document related to the data, to extract related payment grounds related to the target events from a document that describes grounds for insurance payment. Furthermore, this information processing device can also support decision-making in the assessment of insurance payment.

Description

Technical Field

[0001] The present disclosure relates to an information processing apparatus, a support method, and a support program.

Background Art

[0002] When an insurance payment reason occurs, the insurance beneficiary submits necessary documents to the insurance company to claim the insurance payment. Thereafter, a determination of whether the insurance payment can be made by the insurance company (also called payment assessment) is made, and if it is determined that payment is possible, the insurance payment is made to the beneficiary. Therefore, when an event that is considered to be an insurance payment target (for example, illness or injury in the case of medical insurance) occurs, the beneficiary needs to reread the contract terms and the like in which the insurance payment reason is described to confirm whether the event that has occurred corresponds to the insurance payment reason.

[0003] Here, as a document that discloses the prior art related to the insurance payment assessment, for example, the following Patent Document 1 can be cited. Patent Document 1 describes a non-payment target possibility determination device that determines the possibility that an insurance payment claim does not become an insurance payment target using insurance payment claim information from a customer and information described in documents related to the insurance payment claim.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] For many beneficiaries, the aforementioned process of reviewing contract terms and conditions written in technical jargon to confirm whether an event occurred that qualifies as an event for insurance payment is not easy. This is because many beneficiaries do not have sufficient knowledge about insurance, and also because they are often in a mentally or physically unstable state when filing an insurance claim.

[0006] Here, the device for determining the possibility of non-payment described in Patent Document 1 is a device for assisting insurance companies in their insurance payment assessment work when they receive an insurance claim. Therefore, even if this device is used, it is not possible to simplify the insurance claim process. The exemplary purpose of this disclosure is to provide a technology that simplifies the insurance claim process. [Means for solving the problem]

[0007] An information processing device relating to an exemplary aspect of this disclosure includes: an acquisition means for acquiring event information indicating a target event which is an event that may constitute an event that qualifies as an event for payment of insurance benefits; and an extraction means for extracting related payment events related to the target event from a document that describes events for payment of insurance benefits, using an extraction model that has been trained to take a pair of documents and data as input and output a portion of the document that is related to the data.

[0008] Other information processing devices relating to illustrative aspects of this disclosure include acquisition means for acquiring event information indicating a target event which is an event that may constitute an event for payment of insurance benefits, and determination means for determining whether or not the target event constitutes an event for payment using a language model that has been trained on natural language.

[0009] In the exemplary aspects of this disclosure, the support method involves at least one processor performing an acquisition process to acquire event information indicating a target event which is an event that may constitute an event that qualifies as an event for payment of insurance benefits, and an extraction process to extract related payment events related to the target event from a document that describes the events for payment of insurance benefits, using an extraction model that has been trained to take a document and data set as input and output the portion of the document that is related to the data.

[0010] The support program relating to an illustrative aspect of this disclosure causes a computer to function as an acquisition means for acquiring event information indicating a target event, which is an event that may constitute an event that qualifies as an event for payment of insurance benefits, and as an extraction means for extracting related payment events related to the target event from a document that describes the events for payment of insurance benefits, using an extraction model that has been trained to take a pair of documents and data as input and output the portion of the document that is related to the data. [Effects of the Invention]

[0011] One illustrative aspect of this disclosure is that it can facilitate the process of claiming insurance benefits. [Brief explanation of the drawing]

[0012] [Figure 1] This is a block diagram showing the configuration of the information processing device related to this disclosure. [Figure 2] This flowchart shows the flow of support methods related to this disclosure. [Figure 3] This is a block diagram showing the configuration of other information processing devices related to this disclosure. [Figure 4] This figure shows an example of extracting related payment events. [Figure 5] This figure shows an example of how to determine whether a target event falls under a related payment event. [Figure 6] This figure shows an example of a UI (User Interface) screen that accepts correction instructions in natural language. [Figure 7]This figure shows an example of a UI screen that accepts corrections and additions to the description of the target event. [Figure 8] This figure shows an example of a UI screen that accepts input for a description of the history information. [Figure 9] Figure 3 is a flowchart showing the processing flow executed by the information processing device. [Figure 10] This is a block diagram showing the configuration of an information processing device related to a reference example. [Figure 11] This is a block diagram showing the configuration of a computer that functions as an information processing device related to this disclosure. [Modes for carrying out the invention]

[0013] The following are examples of embodiments of the present invention. However, the present invention is not limited to the exemplary embodiments shown below, and various modifications are possible within the scope of the claims. For example, embodiments obtained by appropriately combining some or all of the technologies (things or methods) employed in each of the exemplary embodiments shown below may also be included in the scope of the present invention. Furthermore, embodiments obtained by appropriately omitting some of the technologies employed in each of the exemplary embodiments shown below may also be included in the scope of the present invention. In addition, the effects mentioned in each of the exemplary embodiments shown below are examples of effects that can be expected in that exemplary embodiment and do not define the scope of the present invention. That is, embodiments that do not produce the effects mentioned in each of the exemplary embodiments shown below may also be included in the scope of the present invention.

[0014] [First Exemplary Embodiment] A first exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. This exemplary embodiment is a basic form for each of the exemplary embodiments described below. Note that the scope of application of each technology adopted in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technology adopted in this exemplary embodiment can also be adopted in other exemplary embodiments included in the present disclosure as long as there are no particular technical obstacles. Also, each technology shown in the drawings referred to for explaining this exemplary embodiment can also be adopted in other exemplary embodiments included in the present disclosure as long as there are no particular technical obstacles.

[0015] (Configuration of Information Processing Apparatus 1) The configuration of the information processing apparatus 1 according to this exemplary embodiment will be described with reference to FIG. 1. FIG. 1 is a block diagram showing the configuration of the information processing apparatus 1. As shown in FIG. 1, the information processing apparatus 1 includes an acquisition unit 101 and an extraction unit 102.

[0016] The acquisition unit 101 acquires event information indicating a target event, which is an event that may correspond to the reason for paying an insurance benefit. Note that the above "event" can also be paraphrased as a matter, a fact, or an occurrence, etc. Also, the above "reason for payment" indicates a fact or event that causes an insurance benefit to be paid, and can also be paraphrased as an insurance benefit payment requirement, an insurance benefit payment regulation, an insurance benefit payment condition, etc. Further, the above "insurance benefit" includes not only the insurance benefit paid at the end of an insurance contract (such as a life insurance benefit), but also benefits paid during the continuation of an insurance contract (such as a hospitalization benefit). Also, the insurance underlying an insurance benefit claim is arbitrary. For example, it may be an insurance related to the life or health of an insured person such as life insurance, pension insurance, or medical insurance, or it may be an insurance related to owned property such as automobile insurance or fire insurance.

[0017] Also, the above "event information" may be any information that indicates at least one target event that may correspond to the reason for payment. For example, the acquisition unit 101 may acquire text data in which the insured person has described an event that is considered to have a possibility of receiving an insurance payment as event information. In the following, an example in which the insured person, who is the recipient of the insurance payment, uses the information processing device 1 to consider the request for payment of the insurance payment will be described. However, the person who considers the request for payment of the insurance payment, in other words, the user of the information processing device 1, is arbitrary.

[0018] Also, for example, the acquisition unit 101 may acquire history information indicating a history related to the reason for payment as event information. The history information may be any information that indicates a history that may correspond to the reason for payment. For example, if the target insurance is insurance related to the health of the insured person such as medical insurance, the acquisition unit 101 may acquire history information indicating the insured person's injury and illness history, hospitalization history, outpatient visit history, history of prescribed medications, history of doctor consultations, and history of medical examination results, etc.

[0019] Note that the method of acquiring the event information is arbitrary. Also, the acquisition unit 101 may acquire multiple types of event information. For example, the acquisition unit 101 may acquire both the text data input by the insured person and the history information recorded in a predetermined database (for example, a database that records various types of information related to medical insurance, etc., in association with the personal identification information assigned to each citizen) as event information.

[0020] The extraction unit 102 uses an extraction model that is machine-learned to output a location related to the data in the document when a document-data pair is input, and extracts a payment reason related to the target event indicated in the event information acquired by the acquisition unit 101 from the document in which the payment reason for the insurance payment is described. Since the payment reason extracted by the extraction unit 102 is a payment reason related to the target event, this payment reason will be referred to as a related payment reason below.

[0021] The "extraction model" described above can be any model that can be used to extract relevant payment events from a document describing insurance payment events. For example, a model that divides a document describing payment events into predetermined units of text (e.g., each sentence or each payment event), calculates a score indicating the similarity of the content to the target event for each of the resulting texts, and outputs texts whose calculated scores are above a predetermined threshold can be used as the extraction model. Such a model can be generated, for example, by machine learning using training data that associates the above scores in text and data pairs with the correct data. The data paired with the text is typically text data, but other forms of data such as image data can also be used.

[0022] Furthermore, the "extraction model" described above may be a general-purpose model usable for purposes other than the extraction of relevant payment events, or it may be a general-purpose model that has been fine-tuned for the extraction of relevant payment events. Also, the extraction model may be provided by the information processing device 1, or it may be provided by another device. In the latter case, the extraction unit 102 utilizes the extraction model via the other device that provides the extraction model.

[0023] Furthermore, the "document describing the insurance payment conditions" mentioned above only needs to describe at least some of the payment conditions. For example, the insurance policy, the insurance explanation document, the insurance Q&A collection, a webpage explaining the insurance (e.g., the insured's personal page), or the insurance contract terms and conditions can all be used as documents describing the insurance payment conditions.

[0024] As described above, the information processing device 1 according to this exemplary embodiment is configured to include: an acquisition unit 101 that acquires event information indicating a target event which is an event that may be a cause for payment of insurance benefits; and an extraction unit 102 that uses an extraction model trained on machine learning to take a set of document and data as input and output the portion of the document related to the data, to extract related payment causes related to the target event from a document that describes the causes for payment of insurance benefits.

[0025] According to the above configuration, it is possible to extract relevant payment events related to the target event from a document that describes the events for which insurance benefits are paid. This eliminates the need for the insured to perform the cumbersome task of searching for relevant payment events related to the target event within a document that describes the events for which insurance benefits are paid. Therefore, the information processing device 1 has the effect of simplifying the process of claiming insurance benefits.

[0026] Furthermore, the information processing device 1 can also support decision-making in payment assessment, that is, determining whether or not the event in question qualifies as a payment event. In this way, the information processing device 1 can also contribute to improving the efficiency of payment assessment in insurance companies and the like.

[0027] It is at the discretion of the insured how the extracted relevant payment events are used in insurance claim processing. For example, the information processing device 1 may present the extracted relevant payment events to the insured as reference information. In this case, the insured can decide whether or not to file an insurance claim based on the presented relevant payment events, or file a claim based on the presented relevant payment events. Also, when using the information processing device 1 for payment assessment, the extracted relevant payment events may be presented to the assessor as reference information. Furthermore, for example, as described in Exemplary Embodiment 2 below, the device may automatically determine whether or not a target event falls under a relevant payment event and present the result of that determination.

[0028] (Support Program) The functions of the information processing device 1 described above can also be implemented by a program. The support program according to this exemplary embodiment is a support program for claiming insurance benefits, and causes a computer to function as an acquisition means for acquiring event information indicating a target event which is an event that may be an event that constitutes an event that qualifies as an event for payment of insurance benefits, and as an extraction means for extracting related payment events related to the target event from a document that describes the events that qualify as an event for payment of insurance benefits, using an extraction model that has been trained to take a pair of documents and data as input and output the portion of the document that is related to the data. This support program has the effect of making it easier to claim insurance benefits.

[0029] (Flow of support methods) The flow of the support method according to this exemplary embodiment will be explained with reference to Figure 2. Figure 2 is a flowchart showing the flow of the support method. Note that the entity executing each step in this support method may be a processor provided in the information processing device 1, a processor provided in another device, or the entity executing each step may be a processor provided in a different device.

[0030] In S1 (acquisition process), at least one processor acquires event information indicating a target event that may be an event that triggers the payment of insurance benefits.

[0031] In S2 (extraction process), at least one processor uses an extraction model trained on machine learning to take a document and data pair as input and output the portion of the document that is relevant to the data in question, to extract relevant payment events related to the target event indicated in the event information obtained in S1 from the document that describes the events for which insurance payments are made.

[0032] As described above, the support method according to this exemplary embodiment is a method for supporting insurance claim payments, and employs a configuration in which at least one processor performs an acquisition process to acquire event information indicating a target event which is an event that may be an event that constitutes an event that qualifies as an event for insurance payment, and an extraction process to extract related payment events related to the target event from a document that describes the events that qualify as insurance payments, using an extraction model that has been trained to take a pair of documents and data as input and output the portion of the document that is related to the data. This support method has the effect of making insurance claim payments easier.

[0033] [Second exemplary embodiment] A second exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. Components having the same function as those described in the above-described exemplary embodiment are denoted by the same reference numerals, and their descriptions are omitted as appropriate. The scope of application of each technology adopted in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technology adopted in this exemplary embodiment can also be adopted in other exemplary embodiments included in this disclosure, to the extent that no particular technical problems arise. Furthermore, each technology shown in the drawings referenced to describe this exemplary embodiment can also be adopted in other exemplary embodiments included in this disclosure, to the extent that no particular technical problems arise.

[0034] (Configuration of Information Processing Device 1A) The configuration of the information processing device 1A according to this exemplary embodiment will be described with reference to Figure 3. Figure 3 is a block diagram showing the configuration of the information processing device 1A. The information processing device 1A is a device equipped with a function to support insurance payment claims. The information processing device 1A may be a local device used by individual users, or it may be a server that provides insurance payment claim support services to multiple users.

[0035] As shown in the figure, the information processing device 1A includes a control unit 10A that controls all parts of the information processing device 1A, and a storage unit 11A that stores various data used by the information processing device 1A. The information processing device 1A also includes a communication unit 12A for the information processing device 1A to communicate with other devices, an input unit 13A that receives input to the information processing device 1A, and an output unit 14A for the information processing device 1A to output data. The control unit 10A includes an acquisition unit 101A, an extraction unit 102A, a determination unit 103A, a presentation control unit 104A, a reception unit 105A, and a billing support unit 106A.

[0036] The acquisition unit 101A acquires event information indicating a target event, which is an event that may constitute an event that triggers payment of insurance benefits, similar to the acquisition unit 101 in the exemplary embodiment 1. The acquisition unit 101A may also acquire various other information necessary for claiming insurance benefits other than the event that triggers payment (for example, insured person identification information such as name and date of birth, and the identification number of the insurance policy in question).

[0037] Similar to the extraction unit 102 in Exemplary Embodiment 1, the extraction unit 102A uses an extraction model trained on machine learning to take a document and data pair as input and output the portion of the document related to the data, to extract relevant payment events related to the target event from a document that describes the events for which insurance payments are made. Hereinafter, the extraction model used by the extraction unit 102A will be referred to as extraction model M1.

[0038] The determination unit 103A determines whether the target event indicated in the event information acquired by the acquisition unit 101A corresponds to a payment event, using a language model trained on natural language. More specifically, the determination unit 103A determines whether the target event indicated in the event information acquired by the acquisition unit 101A corresponds to a related payment event extracted by the extraction unit 102A.

[0039] However, the extraction of relevant payment events by the extraction unit 102A is not a mandatory prerequisite for the determination by the determination unit 103A. For example, the determination unit 103A may perform a process to determine whether the target event indicated in the event information acquired by the acquisition unit 101A corresponds to a payment event for each payment event of the insurance. Compared to the case in which relevant payment events are extracted, the number of determinations performed by the determination unit 103A will increase, but even if such a process is adopted, it is still possible to support the claim for payment of insurance benefits.

[0040] Here, machine learning natural language means, more specifically, learning the arrangement of its constituent elements (such as words) in natural language sentences, and the arrangement of sentences in a document. Examples of language models that have learned natural language include BERT (Bidirectional Encoder Representations from Transformers), RoBERTa (Robustly optimized BERT approach), and ELECTRA (Efficiently Learning an Encoder that Classifies Token Replacements Accurately). In the following, the language model used by the determination unit 103A will be referred to as language model M2.

[0041] In this exemplary embodiment, we describe an example in which the language model M2 accepts input in the form of a text prompt written in natural language and outputs a response in natural language. However, the language model M2 may also be a model that can accept input in the form of data other than text data, such as images.

[0042] The language model M2 may be a general-purpose language model that can be used for purposes other than inferring whether the target event constitutes a payment event, or it may be a general-purpose language model that has been fine-tuned for inferring whether the target event constitutes a payment event. Furthermore, the language model M2 may be provided by the information processing device 1A or by another device. In the latter case, the determination unit 103A utilizes the language model M2 via another device that is equipped with the language model M2.

[0043] The presentation control unit 104A presents various information related to supporting insurance claim payments. For example, the presentation control unit 104A presents the determination result of the determination unit 103A. Alternatively, for example, the presentation control unit 104A may present the inference result output by the language model M2. Furthermore, as will be described in detail later, the presentation control unit 104A may present the above inference result along with the basis for that inference. The method of presenting information is arbitrary. For example, the presentation control unit 104A may present the information by having the output unit 14A output the information, or by having the communication unit 12A output the information to another device. Furthermore, the information can be presented in any manner, such as display, printing, voice, or a combination thereof.

[0044] The reception unit 105A receives various instructions related to support for insurance claim payment. For example, the reception unit 105A receives instructions to modify the results of inference by the language model M2. The method of receiving instructions is arbitrary. For example, the reception unit 105A may receive instructions input via the input unit 13A, or it may receive instructions from other devices via the communication unit 12A.

[0045] The claims support unit 106A assists in the claim for payment of insurance benefits. For example, the claims support unit 106A may notify a designated recipient (such as an insurance company) of the relevant event that the determination unit 103A has determined to be a relevant payment event, along with the identification information of the insured and the insurance policy to which the claim is being made, and request payment of insurance benefits.

[0046] Furthermore, since the judgment result of the judgment unit 103A is not necessarily correct, it is preferable to present the relevant event to the insured person and have them confirm its validity before filing a claim. In addition, the support provided by the claim support unit 106A is not limited to automatically filing insurance claims, but is sufficient to reduce the burden on the insured person in filing an insurance claim.

[0047] For example, if the determination unit 103A determines that the relevant payment event applies, the claim support unit 106A may perform the process of obtaining application forms and other documents necessary for claiming insurance benefits from the insurance company, etc. In this case, the claim support unit 106A may send a notification to a designated notification recipient, such as the insurance company's contact point, indicating the identification information of the insured and the insurance policy to which the claim is being made, and requesting the sending of the application forms and other documents.

[0048] Furthermore, for example, if the relevant payment event relates to the insured person's illness or injury and a doctor's certificate is required for the insurance claim, the Claims Support Department 106A may process the acquisition of that certificate. In this case, the Claims Support Department 106A can send the certificate form to the insured person's doctor and request that the doctor fill it out and return it.

[0049] It should be noted that providing the claims support unit 106A is not mandatory. If the claims support unit 106A is not provided, the insured person can consider whether or not to claim insurance benefits by referring to the judgment result of the judgment unit 103A, which is presented by the presentation control unit 104A.

[0050] As described above, the information processing device 1A includes an acquisition unit 101A that acquires event information indicating target events that may constitute events that qualify as grounds for insurance payment, and an extraction unit 102A that uses an extraction model M1, which has been trained to take a document and data set as input and output the portion of the document related to the data, to extract related payment events that are payment events related to the target events from a document that describes the grounds for insurance payment. Therefore, the information processing device 1A has the effect of simplifying the claim for insurance payment, similar to the information processing device 1.

[0051] Furthermore, as described above, the information processing device 1A includes a determination unit 103A that determines whether or not a target event falls under a related payment event using a language model M2 that has been trained on natural language, and a presentation control unit 104A that presents the determination result of the determination unit 103A. As a result, in addition to the effects of the information processing device 1, it is possible to smoothly decide whether or not to file a claim for insurance payment based on the objective determination result of whether or not a target event falls under a related payment event.

[0052] Furthermore, the determination by the determination unit 103A can be omitted. If the determination by the determination unit 103A is omitted, the presentation control unit 104A only needs to present the relevant payment events extracted by the extraction unit 102A. In this case, the insured only needs to compare the presented relevant payment events with the target event to determine whether or not they can make a claim. This eliminates the need for the insured to search for payment events related to the target event from the document that defines the payment events, thereby simplifying the claim for insurance benefits.

[0053] Furthermore, as described above, the information processing device 1A includes an acquisition unit 101A that acquires event information indicating a target event which is an event that may constitute an event that triggers payment of insurance benefits, and a determination unit 103A that determines whether or not the target event constitutes an event that triggers payment using a language model M2 that has been trained on natural language. With this information processing device 1A, an objective determination result can be obtained as to whether or not the target event constitutes an event that triggers payment. Therefore, the effect of simplifying the claim for payment of insurance benefits can be obtained.

[0054] (Example of extracting related payment events) An example of extracting relevant payment events using the information processing device 1A will be explained with reference to Figure 4. Figure 4 is a diagram showing an example of extracting relevant payment events. In the example in Figure 4, the insured person inputs event information 401 into the information processing device 1A. The input event information 401 is acquired by the acquisition unit 101A provided in the information processing device 1A. Note that the person inputting the event information 401 may be someone other than the insured person.

[0055] Event information 401 indicates an event that the insured believes may constitute a payment event, i.e., the target event described above. Furthermore, event information 401 asks whether the target event is eligible for insurance payment, or in other words, whether the event constitutes a payment event. Since the extraction unit 102A extracts relevant descriptions using the extraction model M1, it can handle event information 401 that the subject has freely written in natural language.

[0056] The insured may, for example, input event information 401 via the input unit 13A, or input event information 401 via the communication unit 12A using their own terminal device. The insured may also input event information 401 as text data or as voice data. In the latter case, the acquisition unit 101A can acquire the event information 401 in text format by having the input voice data recognized by the information processing device 1A or other voice recognition device.

[0057] In the example shown in Figure 4, the information processing device 1A (more specifically, the acquisition unit 101A) acquires document 402 from database D, which describes the payment event for the target insurance, that is, the insurance for which the insured is the beneficiary. Hereafter, the above insurance will be referred to as the target insurance. When configuring the system to acquire document 402 from database D, documents describing the payment events for various insurances should be recorded in database D, associated with the identification information of the insurance. Then, the identification information of the target insurance should be input into the information processing device 1A. This allows the acquisition unit 101A to acquire document 402, which describes the payment event for the target insurance, from database D. The method of acquiring document 402 is arbitrary and is not limited to the example described above. For example, the insured may input document 402 along with event information 401 into the information processing device 1A, or document 402 may be stored in the information processing device 1A in advance. Alternatively, for example, a webpage containing the payment conditions for the insurance policy may be loaded into the information processing device 1A as document 402.

[0058] Next, the information processing device 1A (more specifically, the extraction unit 102A) inputs the event information 401 and document 402 acquired by the acquisition unit 101A as described above into the extraction model M1. As a result, the related payment event 403 is output from the extraction model M1.

[0059] Related payment event 403 is a payment event among the payment events stipulated in Document 402 that relates to the target event shown in Event Information 401. Specifically, Related payment event 403 refers to the payment event "(1) Earthquake Compensation Endorsement: We will pay injury insurance benefits for injuries caused by earthquakes" which relates to the target event shown in Event Information 401, "I fractured my left leg when it was trapped under a bookshelf that fell during the earthquake. As a result, I had to go to the hospital for 20 days. Is this covered?".

[0060] As will be explained in detail below, the determination unit 103A determines whether the target event shown in the event information 401 corresponds to the related payment event 403 extracted as described above. As mentioned above, the presentation control unit 104A may also present the extracted related payment event 403 to the insured. In that case, the presentation control unit 104A may extract and present the portion of the related payment event 403 from document 402, or it may present document 402 with the portion of the related payment event 403 highlighted. The highlighting should be done in a manner that allows the highlighted portion and the unhighlighted portion to be distinguished. For example, the presentation control unit 104A may highlight the portion of the related payment event 403 in document 402 by changing the background color of the highlighted portion, changing the font of the text in the highlighted portion, etc.

[0061] (Example of determining whether the event in question qualifies as a related payment event) Figure 5 shows an example of determining whether a target event falls under a related payment event. As described above, this determination is performed by the determination unit 103A. The language model M2 is also used for this determination. Therefore, the determination unit 103A generates a prompt describing the content of the instructions to the language model M2 and inputs it to the language model M2.

[0062] In the example shown in Figure 5, the information processing device 1A (more specifically, the determination unit 103A) inputs prompt 501 to the language model M2. Prompt 501 includes the related payment event 403 shown in Figure 4 and the target event shown in the event information 401 shown in Figure 4, and instructs the device to infer the relationship between the related payment event 403 and the target event.

[0063] More specifically, the target event shown in event information 401 is described in the "Target Event" field in prompt 501. The related payment event 403 is described in the "Payment Event" field in prompt 501. Furthermore, prompt 501 includes the statement, "You are an employee of an insurance company and are conducting insurance claim assessments." While including such a statement is not mandatory, it is expected to improve inference accuracy.

[0064] Furthermore, prompt 501 instructs the system to check whether the event in question qualifies as a payment event. The wording of the prompt can be modified as appropriate to obtain the desired inference result. For example, the determination unit 103A may generate prompts with different inference instructions depending on the event in question, the type of insurance, the language model used, etc.

[0065] Furthermore, prompt 501 includes an "answer format" and instructs the user to answer using this format. By specifying the answer format in this way, it becomes possible to obtain inference results in the desired format. This answer format also includes an item called "basis for inference." By including such an item, the language model M2 can output the basis for the inference along with the inference result. Note that the instruction to output the basis for the inference may also be given by including a sentence such as "Please answer with the basis along with the inference result" in the prompt.

[0066] Furthermore, prompt 501 may include sentences specifying the output conditions. For example, prompt 501 may include sentences such as, "If multiple events are listed, it is necessary to infer whether each event qualifies as a payment event," or "The number of elements included in the response format must match the number of events." This makes it possible to improve the inference accuracy of the language model M2.

[0067] In prompt 501, everything except the content of "Target Event" and "Payment Reason" is standard. For this reason, the parts of prompt 501 other than the content of "Target Event" and "Payment Reason" may be stored as a standard template in the storage unit 11A or the like. This allows the determination unit 103A to generate prompt 501 by inputting the target event shown in the event information 401 acquired by the acquisition unit 101A and the related payment reason 403 extracted by the extraction unit 102A into the above template.

[0068] The inference result 502 shown in Figure 5 is an example of an inference result obtained by inputting prompt 501 into language model M2. Inference result 502 shows the result of inferring the relationship between the relevant payment event and the target event in the format of the response shown in prompt 501. Specifically, inference result 502 shows the inference result that it is "neutral" whether the target event shown in event information 401 corresponds to the relevant payment event or not. In addition, inference result 502 also shows the basis for this inference result.

[0069] Note that while prompt 501 in Figure 5 is a prompt for determining whether a single target event falls under a related payment event 403, it is also possible to use prompts for determining whether each of multiple target events falls under a related payment event 403. Furthermore, if there are multiple related payment events 403, it is also possible to use prompts for determining which of the multiple related payment events 403 the target event falls under, or whether it falls under none of them.

[0070] The format in which the inference results are output can be specified by a prompt. For example, the determination unit 103A may generate a prompt instructing the user to answer with one of two options: "applies to a payment event" or "does not apply to a payment event," or it may generate a prompt instructing the user to answer with one of three options, including "neutral," as shown in the example in Figure 5. Alternatively, the determination unit 103A may generate a prompt instructing the user to output a numerical value (for example, a number between 0 and 1) indicating the likelihood of the event being related to a payment event. In this case, the determination unit 103A only needs to determine that an event is related to a payment event if the numerical value output for a combination of an event and a related payment event is above a predetermined threshold.

[0071] Furthermore, Figure 5 shows an example screen 503 for presenting the inference results 502. The example screen 503 displays the determination result of the determination unit 103A regarding whether each target event corresponds to a related payment event, along with the reason. The presentation control unit 104A can generate and present a screen like the example screen 503 using the determination result of the determination unit 103A and the inference basis shown in the inference result 502. In this way, the presentation control unit 104A may present the determination result of the determination unit 103A to the insured. Alternatively, the presentation control unit 104A may present the inference basis output by the language model M2 to the insured as a basis for judging the appropriateness of the inference result.

[0072] Furthermore, if the output unit 14A has a function to display and output an image, the display control unit 104A may cause the output unit 14A to display the screen. Alternatively, the display control unit 104A may cause the screen to be displayed on an external display device of the information processing device 1A (for example, a display device provided on the terminal device used by the insured) via the communication unit 12A.

[0073] As described above, the determination unit 103A may generate a prompt 501 that includes the relevant payment event and the target event, instructing it to infer the relationship between the relevant payment event and the target event, and then determine whether the target event corresponds to the relevant payment event based on the inference result 502, which is the output obtained by inputting the generated prompt 501 to the language model M2. This makes it possible to appropriately determine whether the target event corresponds to the relevant payment event.

[0074] (Regarding the presentation of inference results and re-inference) The reception unit 105A may accept correction instructions that express the corrections to the inference results in natural language. This will be explained with reference to Figure 6. Figure 6 is a diagram showing an example of a UI screen that accepts correction instructions in natural language. In addition to the example screen 601, which is an example of a UI screen that accepts correction instructions in natural language, Figure 6 also shows a prompt 602 that instructs re-inference based on the correction instructions in natural language, and a re-inference result 603 obtained by inputting the prompt 602 into the language model M2.

[0075] Screen example 601 displays the inference results of language model M2, the event (target event) and payment event (related payment event) that were the subject of the inference, and prompts the user to check for any errors or points of interest in the inference results. If there are any errors, the user is asked to input the details and instruct the system to re-infer. Screen example 601 also displays a text box for entering corrections and a button (software key) for instructing the system to re-infer.

[0076] The presentation control unit 104A can allow the insured person to confirm the inference results by presenting a UI screen such as the example screen 601. The reception unit 105A can also accept corrections to the inference results and input of supplementary information related to the inference via a UI screen such as the example screen 601.

[0077] In the example screen 601, when a correction or supplementary information is entered into the text box and the button to instruct re-inference is operated, the receiving unit 105A receives the entered correction as a correction instruction. For example, the receiving unit 105A may accept a correction instruction that describes in natural language why the inference result is incorrect. Alternatively, for example, the receiving unit 105A may accept a correction instruction that describes in natural language whether or not the subject event that was the subject of the inference falls under a related payment event. Alternatively, for example, the receiving unit 105A may accept as a correction instruction supplementary information that should be considered in the inference of whether or not the subject event falls under a related payment event.

[0078] Then, the determination unit 103A re-determines whether the target event falls under the relevant payment event based on the correction instruction received by the reception unit 105A. For example, the determination unit 103A may generate the prompt 602 shown in Figure 6 and input it to the language model M2, and then re-determine whether the target event falls under the relevant payment event based on the re-inference result 603 output from the language model M2.

[0079] Prompt 602 contains the text of the correction instruction received by the reception unit 105A and is a prompt that instructs the system to infer the relationship between the target event and the related payment event based on the correction instruction. Prompt 602 also includes the sentence, "However, do not output any words that are not included in the response format." Thus, even in prompts that instruct re-inference, sentences specifying output conditions may be included. This prevents unnecessary content (for example, sentences such as "Apologies. I will re-output." or "Content has been updated") from being output from the language model M2, thus preventing any problems with inputting information into insurance input forms, etc. Prompt 602 can also be generated using a predetermined template, similar to prompt 501 shown in Figure 5. In addition, prompt 602 may also include descriptions of the target event, related payment event, and response format, similar to prompt 501.

[0080] In the re-inference result 603 shown in Figure 6, the inference result regarding the relationship between the target event and the related payment event has changed from that shown on the UI screen 601, reflecting the content of the "comment," or correction instruction, shown in prompt 602. Specifically, in the re-inference result 603, the inference result regarding whether the target event, "I missed a step on the stairs at home and fractured a bone in my right foot, requiring hospital visits for about a month," constitutes a related payment event has changed to "Does not constitute a payment event." Furthermore, the basis for this inference has changed to the sentence, "Accidents occurring under the influence of alcohol are not covered by compensation." In this way, by performing re-inference using prompt 602, which includes the natural language sentence received as a correction instruction, it is possible to output a re-inference result 603 that reflects the content of the correction instruction.

[0081] As described above, the determination unit 103A may generate a prompt (for example, prompt 501 in Figure 5) that instructs the system to output the reasoning result regarding the relationship between the relevant payment event and the target event, along with the basis for that reasoning. The presentation control unit 104A may then present the reasoning result output by the language model M2 along with the basis for that reasoning. This provides the effect that, in addition to the effects of the information processing device 1, the system can consider the appropriateness of the reasoning result of the language model M2 using the basis for that reasoning as a basis for judgment. Alternatively, the presentation control unit 104A may present the determination result of the determination unit 103A along with the above-mentioned basis for that reasoning, instead of the reasoning result output by the language model M2.

[0082] Furthermore, as described above, the information processing device 1A is equipped with a receiving unit 105A that receives correction instructions for the inference results. The determination unit 103A then re-determines whether the target event falls under the relevant payment event based on the correction instructions received by the receiving unit 105A. This provides the effect of correcting errors in the inference results, in addition to the effects achieved by the information processing device 1, so that the insured can correctly file a payment claim. Note that the above-mentioned "correction instructions" include not only instructions to correct the inference results, but also instructions to include supplementary information that should be considered in the inference and to re-infer based on that supplementary information.

[0083] Furthermore, as described above, the reception unit 105A may accept correction instructions expressed in natural language. In this case, the determination unit 103A generates a prompt (for example, prompt 602 in Figure 6) that includes the correction instructions received by the reception unit 105A and instructs the determination unit 103A to infer the relationship between the target event and the related payment event based on the correction instructions. Based on the output obtained by inputting the generated prompt into the language model M2, the determination unit 103A re-determines whether the target event falls under the related payment event. This provides the effect of enabling appropriate re-determination by taking into account the intent of the correction instructions, in addition to the effects performed by the information processing device 1.

[0084] One example of how the intent behind correction instructions can be understood is the absorption of variations in wording. For example, suppose the inference result is that the subject event, as in the example in Figure 6, "I missed a step on the stairs at home and fractured a bone in my right foot, requiring hospital visits for about a month," falls under the related payment event, "We will pay insurance benefits if the insured is injured due to a 'sudden and accidental external event'." In this case, even if the insured enters a sentence different from, for example, "I had been drinking quite a bit just before, is that okay?" as in the example in Figure 6, such as "Is it okay if I've been drinking?" or "I had two or three beers at the time," the re-inference result can be determined to be "Does not fall under the payment event."

[0085] Furthermore, accepting correction instructions expressed in natural language has the advantage of allowing the content of the correction instructions to be reflected in other inferences. For example, suppose the inference result that the target event "I was diagnosed with diabetes at X Clinic and received treatment there for a while" corresponds to the related payment event "We will pay you benefits for hospitalization due to lifestyle-related diseases" is corrected by the input of the correction instruction "X Clinic does not have inpatient facilities." In this case, the judgment unit 103A can not only correct the inference result for the target event to "Does not correspond to a payment event," but can also update other inference results based on the fact that X Clinic does not have inpatient facilities.

[0086] (Reuse of correction instructions) The content of the correction instructions described above may be recorded in the memory unit 11A or an external database, etc., and used for subsequent inferences. For example, the comment of the correction instruction, "X Clinic does not have inpatient facilities," may be recorded, and this comment may be used for inferences regarding other target events. This will allow subsequent inferences (which may be limited to inferences regarding the same insured person, or may be reused for inferences regarding other insured persons) to obtain inference results that take into account that X Clinic does not have inpatient facilities.

[0087] Furthermore, the presentation control unit 104A may present the recorded correction instructions to the insured, and the reception unit 105A may accept corrections, deletions, and additions to the correction instructions. This allows the content of the correction instructions to be reflected in subsequent inferences without requiring retraining of the language model M2. The correction instructions can be reused in the same way as when they are used for the first time, by including the correction instructions in the prompt. In other words, the determination unit 103A generates a prompt that includes the recorded correction instructions and instructs the language model M2 to infer the relationship between the target event and the related payment event based on the correction instructions, and then inputs the generated prompt to the language model M2 to output an inference result indicating whether or not the target event falls under the related payment event.

[0088] (Regarding iterative reasoning) Since the language model M2 is a probabilistic model, even if the exact same prompt is entered, the inference results in multiple inferences may differ. Furthermore, it has been found that inference results that differ from the facts tend not to be output repeatedly. For this reason, the determination unit 103A may perform the process of inputting a prompt to the language model M2 and outputting an inference result multiple times. In this case, the determination unit 103A may determine that events with large variations in inference results do not fall under the category of related payment events. For example, the determination unit 103A may calculate a score for each event indicating the magnitude of the variation in inference results (for example, the proportion of inference results that differ in content from other inference results out of all inference results), and determine that events whose calculated score exceeds a predetermined threshold do not fall under the category of related payment events.

[0089] (Regarding revisions and additions to the description of the target event) As shown in the example in Figure 5, if event information with insufficient explanation of the target event is input, it may be impossible to determine whether or not the target event qualifies for payment, or an incorrect determination result may be output. For this reason, the information processing device 1A may prompt the user to correct or supplement the explanation of the target event after the determination unit 103A has made a determination. This will be explained with reference to Figure 7. Figure 7 shows an example of a UI screen that accepts corrections and supplements to the explanation of the target event.

[0090] In the example screen 701 shown in Figure 7, the judgment result of the judgment unit 103A is displayed along with its basis (more precisely, the basis for the inference of the language model M2 that formed the basis of the judgment result). In addition, the example screen 701 displays text prompting the user to enter any additional information or corrections. The example screen 701 also displays a text box for entering additional information or corrections, and a "Confirm eligibility for payment" button (software key) to allow the user to confirm whether they are eligible for payment. When the presentation control unit 104A presents the judgment result of the judgment unit 103A, it is possible to extract information from the insured person that is necessary to obtain a more appropriate judgment result by presenting a UI screen like the example screen 701.

[0091] In the example screen 701, if the entered event information displayed in the text box is modified or additional information is entered, and the "Confirm payment eligibility" button is operated, the acquisition unit 101A acquires the entered content as new event information. Subsequently, as in the case where event information was acquired earlier, the extraction unit 102A extracts related payment events, and the determination unit 103A makes a determination, and the determination result of whether the target event falls under a payment event is presented.

[0092] For example, as shown in the example in Figure 7, suppose that information regarding the application status of the earthquake compensation rider is added without changing the already entered event information. In this case, the extraction unit 102A extracts the portion describing the earthquake compensation rider as a related payment event, as before. The determination unit 103A also generates a prompt instructing the language model M2 to infer whether the target event falls under a related payment event, including the added information, and inputs it. Here, since the added information suggests that an earthquake compensation rider contract has been made, the inference result output by the language model M2 will indicate that the target event falls under a related payment event. For this reason, the determination unit 103A determines that the target event falls under a related payment event and causes the presentation control unit 104A to present the determination result.

[0093] (Regarding the use of historical information) The event information acquired by the acquisition unit 101A may be matters entered by the insured that they believe may constitute a payment event, or it may be the insured's medical history information, or it may be both. In this case, if the acquisition unit 101A acquires both the information entered by the insured and the history information as event information, there may be matters that are not shown in the information entered by the insured but are shown in the history information that may constitute a payment event.

[0094] Therefore, the acquisition unit 101A may acquire target events from the information entered by the insured, as well as from the history information, and then perform a process of matching these target events. If the presentation control unit 104A detects a target event that was not acquired from the information entered by the insured but was acquired from the history information through the above process, it may present that target event to the insured and prompt them to confirm whether it is eligible for payment. This makes it possible to claim insurance payments for events that the insured was not aware were eligible for insurance payment.

[0095] Alternatively, the above-mentioned matching process may be omitted, and the historical information may be used directly as event information to extract related payment events. In this case, the extraction unit 102A may input the event shown in the historical information and the document describing the insurance payment event into the extraction model M1, and perform the process of extracting related payment events associated with that event for each event shown in the historical information.

[0096] Furthermore, the historical information acquired by the acquisition unit 101A may not contain sufficient information to determine whether each event shown in the historical information constitutes an event for insurance payment. For this reason, the presentation control unit 104A may present the target events acquired from the historical information to the insured and prompt them to input information to determine whether or not they constitute an event for payment. This will be explained with reference to Figure 8.

[0097] Figure 8 shows an example of a UI screen that accepts input for a description of historical information. In the example screen 801 shown in Figure 8, a message indicating that historical information potentially related to the payment reason has been detected is displayed, along with the detected historical information. Specifically, this historical information indicates that the patient was hospitalized at Y Hospital from September 1st to 5th, 2021.

[0098] Furthermore, Screen Example 801 displays text prompting the user to enter the reason for hospitalization and press the "Check if eligible for payment" button. In addition, Screen Example 801 also displays a text box for entering the reason for hospitalization and a "Check if eligible for payment" button (software key) to instruct the system to determine whether the information shown in the history data constitutes a payment event, taking the entered reason into consideration. The reception unit 105A can accept the input of the reason for hospitalization through such a UI screen.

[0099] In example screen 801, when the reason for hospitalization is entered in the text box and the "Confirm eligibility for payment" button is pressed, the extraction unit 102A extracts the relevant payment events. In this case, the extraction unit 102A treats the detected history information and the entered reason for hospitalization as a single target event and extracts the relevant payment events associated with that target event. For example, if the entered reason for hospitalization is "heart failure," the extraction unit 102A inputs the sentence "2021 / 9 / 1-5: Admitted to Y Hospital, Reason for hospitalization: Heart failure" as the target event into the extraction model M1 and extracts the relevant payment events. Subsequently, the determination unit 103A performs a determination and presents the determination result of whether or not the items shown in the history information correspond to the relevant payment events.

[0100] In the example shown in Figure 8, the detected history information indicates a hospitalization history. Therefore, the reception unit 105A accepts input for the reason for hospitalization. Thus, the reception unit 105A only needs to accept input for additional information corresponding to the content of the history information, in other words, for explanatory text that explains the history information. For example, if history information indicating a drug prescription history is detected, the reception unit 105A only needs to accept input for explanatory text about the prescribed drug.

[0101] As described above, the information processing device 1A includes a presentation control unit 104A that presents the insured's history information regarding payment events, and a reception unit 105A that receives input of an explanatory text describing the presented history information. The extraction unit 102A treats the set of the history information and the explanatory text as a single target event and extracts related payment events associated with that target event. This provides the effect that, in addition to the effects of the information processing device 1, it is possible to extract appropriate related payment events that take into account the input explanatory text. Furthermore, the determination unit 103A may use the language model M2 to determine whether the target event, including the history information and the explanatory text, corresponds to an appropriate payment event. This provides the effect that it is possible to obtain an appropriate determination result that takes into account the input explanatory text.

[0102] (Process flow) The processing flow performed by the information processing device 1A will be explained with reference to Figure 9. Figure 9 is a flowchart showing the processing flow performed by the information processing device 1A. The flowchart in Figure 9 includes each process of the support method according to this exemplary embodiment.

[0103] In S11 (acquisition process), the acquisition unit 101A acquires event information indicating a target event that may be an event that constitutes an event for which insurance benefits are payable. For example, the acquisition unit 101A may acquire event information that the insured person inputs to the information processing device 1A.

[0104] In S12 (extraction process), the extraction unit 102A uses an extraction model M1, which has been trained to take a document and data pair as input and output the portion of the document related to the data, to extract related payment events, which are payment events related to the target event indicated in the event information obtained in S11, from the document describing the event reasons for insurance payment. Note that the document describing the event reasons for insurance payment may be input along with the event information in S11, or it may be obtained from a predetermined database as shown in the example in Figure 4.

[0105] In S13, the determination unit 103A generates a prompt to be input to the language model M2. Specifically, the determination unit 103A generates a prompt that includes the relevant payment event extracted in S12 and the target event shown in the event information acquired in S11, and instructs the determination unit 103A to infer the relationship between the relevant payment event and the target event, in other words, whether the target event corresponds to the relevant payment event.

[0106] In S14, the determination unit 103A inputs the prompt generated in S13 into the language model M2 to infer the relationship between the relevant payment event and the target event. In S15, the presentation control unit 104A presents the inference result from S14 to the insured. In S16, the reception unit 105A determines whether there is a request for correction to the inference result presented in S15. If the result in S16 is YES, the process proceeds to S17; if the result in S16 is NO, the process proceeds to S19.

[0107] In S15, the presentation control unit 104A may present the inference result by displaying a UI screen, such as the example screen 601 in Figure 6. In that case, the reception unit 105A may receive correction instructions via the UI screen. The process of presenting the inference result may be omitted. In that case, the process proceeds to S19 after S14.

[0108] In S17, the reception unit 105A records the contents of the received correction instruction in the storage unit 11A or an external database, etc. The timing of recording the contents of the correction instruction is arbitrary. For example, the reception unit 105A may record the contents of the correction instruction after the determination described later has been completed, or after the insurance payment claim has been completed.

[0109] In S18, the determination unit 103A generates a prompt that reflects the content of the correction instruction received by the reception unit 105A, specifically a prompt that includes the correction instruction and instructs the system to infer the relationship between the target event and the related payment event based on the correction instruction. After this, the process returns to S14, where the determination unit 103A inputs the newly generated prompt into the language model M2 to infer the relationship between the target event and the related payment event.

[0110] In S19, the determination unit 103A determines, based on the inference result of the language model M2 (or the latest inference result if multiple inferences were performed), whether the target event indicated in the event information acquired in S11 corresponds to the related payment event extracted in S12. Then, in S20, the presentation control unit 104A presents the determination result from S19 to the insured.

[0111] In S21, the billing support unit 106A determines whether or not to issue a payment request. If the result in S21 is YES, the process proceeds to S22. On the other hand, if the result in S21 is NO, the process returns to S11.

[0112] For example, if the determination unit 103A determines in S19 that the relevant payment event applies, the presentation control unit 104A may present the determination result along with the option of whether or not to make a payment request. If the option to make a payment request is selected from the presented options, the billing support unit 106A should then determine in S21 to make a payment request.

[0113] On the other hand, if the determination unit 103A determines in S19 that the relevant payment event does not apply, or if it is unable to determine whether or not the relevant payment event applies, the presentation control unit 104A may present the determination result in S20 by displaying a UI screen such as the example screen 701 in Figure 7. In this case, in S21, the billing support unit 106A may determine not to issue a payment request.

[0114] In S11, which follows a transition from S21, the acquisition unit 101A acquires new event information. For example, if a UI screen like the example screen 701 in Figure 7 is displayed in S20, the acquisition unit 101A only needs to acquire the information entered via that UI screen as new event information.

[0115] In S22, the claims support unit 106A makes a payment claim for the event that was determined to be eligible for payment in S21. For example, the claims support unit 106A may generate a notice requesting payment of insurance benefits on the grounds that the event in question falls under the relevant payment event, and send the generated notice to a designated recipient (such as an insurance company). This completes the process shown in Figure 10.

[0116] Furthermore, as mentioned above, the inference results of the language model M2 are not always correct. For this reason, the billing support unit 106A may transmit the event information obtained in S11 and the inference results in S14 as reference information along with the payment request notification. Also, if the basis for the inference is output along with the inference results, the billing support unit 106A may transmit the basis for the inference as reference information as well. This makes it possible for the payment assessor to appropriately determine whether the event in question constitutes a payment event by considering this reference information.

[0117] [Reference example 1] Figure 10 is a block diagram showing the configuration of the information processing device 1B according to this reference example. As shown in the figure, the information processing device 1B includes an acquisition unit 101B and a determination unit 103B.

[0118] The acquisition unit 101B acquires event information indicating a target event that may be an event that constitutes an event that triggers payment of insurance benefits, similar to the acquisition unit 101A in the exemplary embodiment 2.

[0119] Similar to the determination unit 103A in the exemplary embodiment 2, the determination unit 103B determines whether the target event indicated in the event information acquired by the acquisition unit 101B constitutes a payment event, using a language model trained on natural language.

[0120] The above payment events may be extracted using an extraction model from a document describing payment events (i.e., related payment events), as in exemplary embodiment 1 or 2, or they may not be extracted using an extraction model. For example, the acquisition unit 101B may acquire the payment events corresponding to the target event, in other words, the payment events for which the determination of whether or not the target event applies, by having the insured input them. Alternatively, for example, the determination unit 103B may extract a part of a document describing the insurance payment events (for example, a sentence or paragraph with a coherent content) by analyzing the document and use that part to perform the above determination.

[0121] As described above, the information processing device 1B includes an acquisition unit 101B that acquires event information indicating target events, which are events that may constitute an event that triggers payment of insurance benefits, and a determination unit 103B that uses a language model trained on natural language to determine whether or not the target events indicated in the event information acquired by the acquisition unit 101B constitute an event that triggers payment. With this information processing device 1B, an objective determination result can be obtained regarding whether or not a target event constitutes an event that triggers payment. Therefore, the effect of simplifying the claim for payment of insurance benefits can be obtained.

[0122] Furthermore, how the determination result of the determination unit 103B is used to facilitate payment claims is at the discretion of the user. For example, the information processing device 1B may be equipped with a presentation control unit 104A similar to the information processing device 1A in the exemplary embodiment 2. In this case, the presentation control unit 104A can be made to present the determination result of the determination unit 103B. This allows the insured person to consider making a payment claim by referring to the presented determination result.

[0123] Furthermore, for example, the information processing device 1B may be equipped with a claim support unit 106A similar to the information processing device 1A in the exemplary embodiment 2. In this case, the claim support unit 106A can be instructed to make a claim for payment of insurance benefits for the target event that the determination unit 103B has determined to be a payment event.

[0124] (Support Program) The functions of the information processing device 1B described above can also be implemented by a program. The support program in this reference example is a support program for insurance payment claims, and the computer functions as an acquisition means for acquiring event information indicating target events that may constitute events that qualify as grounds for insurance payment, and as a determination means for determining whether the target events indicated in the event information acquired by the acquisition means qualify as grounds for payment, using a language model that has been trained on natural language. This support program has the effect of simplifying the process of claiming insurance payments.

[0125] (Support method) The support method described in this reference example is a method for supporting insurance claim payments, in which at least one processor performs an acquisition process to acquire event information indicating a target event which is an event that may constitute an event that warrants payment of insurance benefits, and a determination process to determine whether or not the target event indicated in the event information acquired in the acquisition process constitutes an event that warrants payment, using a language model that has been trained on natural language. This support method has the effect of making insurance claim payments easier.

[0126] [Reference example 2] In the exemplary embodiments described above, information processing devices 1 and 1A were described that acquire event information indicating a target event which is an event that may constitute an event that triggers payment of insurance benefits, and extract related payment events related to the target event from a document that describes the events that trigger insurance benefits.

[0127] Furthermore, the above-mentioned reference example describes an information processing device 1B that acquires event information indicating a target event which is an event that may constitute an event that triggers payment of insurance benefits, and determines whether or not the target event constitutes an event that triggers payment.

[0128] These information processing devices 1, 1A, and 1B can be used not only to support insurance claim processing, but also to determine whether any event meets any requirements. For example, when preparing a subsidy application, the applicant determines whether they meet the subsidy eligibility requirements and then prepares an application document demonstrating that they do.

[0129] In this process, the applicant can input potentially relevant information into Information Processing Device 1 or 1A instead of event information, and have Information Processing Device 1 refer to the document containing the eligibility requirements instead of the document containing the payment reasons. This allows the device to extract the requirements from the document containing the eligibility requirements that are relevant to the information entered by the applicant. By having Information Processing Device 1 or 1A present the extracted requirements, the applicant can quickly determine whether they meet the requirements and efficiently proceed with preparing the application.

[0130] Furthermore, the applicant may input items that may qualify as eligibility requirements into the information processing device 1B instead of event information, and may also have the information processing device 1B refer to the eligibility requirements for the subsidy. This allows the information processing device 1B to output a result of whether or not the items entered by the applicant meet the eligibility requirements. The applicant can then use the presented result as a reference to efficiently prepare the application form.

[0131] In addition, information processing devices 1, 1A, and 1B can also be used, for example, to create product manuals that conform to the specifications, or to create documents that comply with laws, regulations, and other provisions. Furthermore, information processing devices 1, 1A, and 1B can also be used to design devices and systems that conform to standards. For example, by using information processing device 1 or 1A, communication-related descriptions from standards can be extracted when considering communication specifications for devices and systems. Also, by using information processing devices 1, 1A, or 1B, it is possible to easily verify whether the design specifications conform to standards.

[0132] [Variation] The entities executing each process described in the above-described exemplary embodiments and reference examples are arbitrary and not limited to the examples given above. For example, a system having the same functions as information processing devices 1, 1A, and 1B can be constructed using multiple devices that can communicate with each other. Furthermore, the entities executing each process shown in the flowchart in Figure 9 may be a single device (which can also be called a processor) or multiple devices (which can also be called processors).

[0133] [Examples of implementation using software] Some or all of the functions of the information processing devices 1, 1A, and 1B (hereinafter also referred to as "the above devices") may be implemented by hardware such as integrated circuits (IC chips) or by software.

[0134] In the latter case, each of the above devices is implemented, for example, by a computer that executes instructions for a program, which is software that realizes each function. An example of such a computer (hereinafter referred to as computer C) is shown in Figure 11. Figure 11 is a block diagram showing the hardware configuration of computer C, which functions as each of the above devices.

[0135] Computer C comprises at least one processor C1 and at least one memory C2. Memory C2 stores a program (support program) P that causes computer C to operate as each of the above-mentioned devices. In computer C, processor C1 reads program P from memory C2 and executes it, thereby realizing each of the above-mentioned devices.

[0136] For processor C1, for example, a CPU (Central Processing Unit), GPU (Graphic Processing Unit), DSP (Digital Signal Processor), MPU (Micro Processing Unit), FPU (Floating Point Number Processing Unit), PPU (Physics Processing Unit), TPU (Tensor Processing Unit), quantum processor, microcontroller, or a combination thereof can be used. For memory C2, for example, flash memory, HDD (Hard Disk Drive), SSD (Solid State Drive), or a combination thereof can be used.

[0137] Computer C may also be equipped with RAM (Random Access Memory) for loading program P at runtime and for temporarily storing various data. Furthermore, computer C may be equipped with communication interfaces for sending and receiving data with other devices. Additionally, computer C may be equipped with input / output interfaces for connecting input / output devices such as keyboards, mice, displays, and printers.

[0138] Furthermore, program P can be recorded on a non-temporary, tangible recording medium M that is readable by computer C. Such a recording medium M could be, for example, tape, disk, card, semiconductor memory, or programmable logic circuitry. Computer C can acquire program P via such a recording medium M. Program P can also be transmitted via a transmission medium. Such a transmission medium could be, for example, a communication network or broadcast waves. Computer C can also acquire program P via such a transmission medium.

[0139] Furthermore, each of the above functions of each of the above devices may be implemented by a single processor in a single computer, by multiple processors in a single computer working together, or by multiple processors in each of multiple computers working together. In addition, the programs for implementing each of the above functions in each of the above devices may be stored in a single memory in a single computer, distributed and stored in multiple memories in a single computer, or distributed and stored in multiple memories in each of multiple computers.

[0140] [Additional Notes] This disclosure includes the technologies described in the following appendices. However, the present invention is not limited to the technologies described in the following appendices, and various modifications are possible within the scope of the claims.

[0141] (Note A1) An information processing device comprising: an acquisition means for acquiring event information indicating a target event which is an event that may constitute an event that qualifies as an event for payment of insurance benefits; and an extraction means for extracting related payment events related to the target event from a document that describes events that qualify as an event for payment of insurance benefits, using an extraction model that has been trained by machine learning to take a pair of documents and data as input and output the portion of the document that is related to the data.

[0142] (Appendix A2) The information processing apparatus according to Appendix A1, comprising: determination means for determining whether the aforementioned target event falls under the aforementioned related payment event using a language model that has been trained on natural language; and presentation control means for presenting the determination result of the determination means.

[0143] (Note A3) The information processing device according to Appendix A2, wherein the determination means includes the related payment event and the target event, generates a prompt instructing the system to infer the relationship between the related payment event and the target event, and determines whether the target event corresponds to the related payment event based on the output obtained by inputting the generated prompt to the language model.

[0144] (Note A4) The information processing apparatus according to Appendix A3, wherein the determination means generates a prompt instructing the determination means to output the basis for the inference along with the inference result of the relationship between the related payment event and the target event, and the presentation control means presents the inference result output by the language model or the determination result of the determination means along with the basis.

[0145] (Note A5) The information processing device described in Appendix A4, comprising a receiving means for receiving instructions for correction of the inference result, wherein the determination means re-determines whether the target event falls under the related payment event based on the correction instructions received by the receiving means.

[0146] (Note A6) The information processing device described in Appendix A5, wherein the receiving means receives a correction instruction expressing the content of the correction in natural language, the determination means generates a prompt that includes the correction instruction received by the receiving means and instructs the determination means to infer the relationship between the target event and the related payment event based on the correction instruction, and inputs the generated prompt into the language model and re-determines whether the target event corresponds to the related payment event based on the output obtained.

[0147] (Note A7) An information processing device according to any one of Appendix A1 to A6, comprising: a presentation control means for presenting historical information showing a history related to an insurance payment event; and a reception means for receiving input of an explanatory text describing the presented historical information, wherein the extraction means extracts related payment events related to the target event, treating the set of the historical information and the explanatory text as a single target event.

[0148] (Note A8) An information processing device comprising: an acquisition means for acquiring event information indicating a target event which is an event that may constitute an event that triggers payment of insurance benefits; and a determination means for determining whether or not the target event constitutes an event that triggers payment using a language model that has been trained on natural language.

[0149] (Note B1) A support method that performs, at least one processor, an acquisition process to acquire event information indicating a target event which is an event that may constitute an event that qualifies as an event for payment of insurance benefits, and an extraction process to extract related payment events related to the target event from a document that describes events that qualify as events for payment of insurance benefits, using an extraction model that has been trained to take a pair of documents and data as input and output the portion of the document that is related to the data.

[0150] (Note B2) The support method according to Appendix B1, comprising: a determination process in which at least one processor determines whether the target event falls under the related payment event using a language model that has been trained on natural language; and a presentation control process in which at least one processor presents the determination result of the determination process.

[0151] (Note B3) The support method according to Appendix B2, wherein in the determination process, the at least one processor generates a prompt that includes the related payment event and the target event and instructs the processor to infer the relationship between the related payment event and the target event, and determines whether the target event corresponds to the related payment event based on the output obtained by inputting the generated prompt to the language model.

[0152] (Note B4) The support method according to Appendix B3, wherein in the determination process, the at least one processor generates a prompt instructing the system to output the basis for the inference along with the inference result of the relationship between the related payment event and the target event, and the at least one processor presents the inference result output by the language model or the determination result of the determination process along with the basis.

[0153] (Note B5) The support method according to Appendix B4, wherein the at least one processor includes an acceptance process for receiving a correction instruction for the inference result, and the at least one processor re-determines whether the target event falls under the related payment event based on the correction instruction received in the acceptance process.

[0154] (Note B6) The support method described in Appendix B5, wherein in the reception process, at least one processor receives a correction instruction expressing the content of the correction in natural language, the at least one processor generates a prompt that includes the correction instruction received in the reception process and instructs the system to infer the relationship between the target event and the related payment event based on the correction instruction, and inputs the generated prompt into the language model to re-determine whether the target event falls under the related payment event based on the output obtained.

[0155] (Note B7) The support method according to any one of Appendix B1 to B6, comprising: a presentation control process in which at least one processor presents historical information showing a history related to an insurance payment event; and a reception process in which at least one processor accepts input of an explanatory text describing the presented historical information, wherein in the extraction process, the at least one processor extracts related payment events associated with the target event, treating the pair of the historical information and the explanatory text as a single target event.

[0156] (Note B8) A support method comprising: an acquisition process in which at least one processor acquires event information indicating a target event which is an event that may constitute an event that warrants payment of insurance benefits; and a determination process in which the at least one processor determines whether or not the target event constitutes an event that warrants payment using a language model that has been trained on natural language.

[0157] (Note C1) A support program that enables a computer to function as an acquisition means for acquiring event information indicating a target event, which is an event that may constitute an event that triggers the payment of insurance benefits, and as an extraction means for extracting related payment events related to the target event from a document that describes the events that trigger the payment of insurance benefits, using an extraction model that has been trained to take a pair of documents and data as input and output the portion of the document that is related to the data.

[0158] (Note C2) The support program described in Appendix C1 causes the computer to function as a determination means for determining whether the target event falls under the related payment event using a language model that has been trained on natural language, and a presentation control means for presenting the determination result of the determination means.

[0159] (Note C3) The support program described in Appendix C2 includes the relevant payment event and the target event, generates a prompt instructing the system to infer the relationship between the relevant payment event and the target event, and determines whether the target event corresponds to the relevant payment event based on the output obtained by inputting the generated prompt to the language model.

[0160] (Note C4) The support program as described in Appendix C3, wherein the determination means generates a prompt instructing the determination means to output the basis for the inference along with the inference result of the relationship between the related payment event and the target event, and the presentation control means presents the inference result output by the language model or the determination result of the determination means along with the basis.

[0161] (Note C5) The support program described in Appendix C4, wherein the computer functions as a receiving means for receiving instructions to correct the inference result, and the determination means re-determines whether the target event falls under the related payment event based on the correction instructions received by the receiving means.

[0162] (Appendix C6) The support program described in Appendix C5, wherein the receiving means receives a correction instruction expressing the content of the correction in natural language, the determination means generates a prompt that includes the correction instruction received by the receiving means and instructs the determination means to infer the relationship between the target event and the related payment event based on the correction instruction, and the determination means re-determines whether the target event falls under the related payment event based on the output obtained by inputting the generated prompt into the language model.

[0163] (Note C7) The computer functions as a presentation control means for presenting historical information showing a history related to an insurance payment event, and a reception means for receiving input of an explanatory text describing the presented historical information, and the extraction means is a support program as described in any of Appendix C1 to C6 for extracting related payment events related to a target event, with the set of the historical information and the explanatory text being treated as a single target event.

[0164] (Note C8) A support program that causes the aforementioned computer to function as an acquisition means for acquiring event information indicating a target event which is an event that may constitute an event that triggers payment of insurance benefits, and a determination means for determining whether or not the target event constitutes an event that triggers payment using a language model that has been trained on natural language.

[0165] (Note D1) An information processing device comprising at least one processor, the at least one processor performing an acquisition process to acquire event information indicating a target event which is an event that may constitute an event that constitutes

[0166] The information processing device may also include memory. Furthermore, the memory may store a program that causes at least one processor to execute each of the aforementioned processes.

[0167] (Note D2) The information processing apparatus described in Appendix D1, wherein the at least one processor performs a determination process that determines whether the target event falls under the related payment event using a language model that has been trained on natural language, and a presentation control process that presents the determination result of the determination process.

[0168] (Note D3) The information processing apparatus according to Appendix D2, wherein in the determination process, the at least one processor generates a prompt that includes the related payment event and the target event and instructs the processor to infer the relationship between the related payment event and the target event, and determines whether the target event corresponds to the related payment event based on the output obtained by inputting the generated prompt to the language model.

[0169] (Note D4) The information processing apparatus according to Appendix D3, wherein in the determination process, the at least one processor generates a prompt instructing the output of the basis for the inference along with the inference result of the relationship between the related payment event and the target event, and the at least one processor presents the inference result output by the language model or the determination result of the determination process along with the basis.

[0170] (Note D5) The information processing apparatus according to Appendix D4, wherein the at least one processor performs an acceptance process to receive a correction instruction for the inference result, and the at least one processor re-determines whether the target event falls under the related payment event based on the correction instruction received in the acceptance process.

[0171] (Note D6) The information processing apparatus according to Appendix D5, wherein in the reception process, at least one processor receives a correction instruction expressing the content of the correction in natural language, the at least one processor generates a prompt that includes the correction instruction received in the reception process and instructs the processor to infer the relationship between the target event and the related payment event based on the correction instruction, and inputs the generated prompt to the language model and re-determines whether the target event corresponds to the related payment event based on the output obtained.

[0172] (Note D7) The information processing device according to any one of the appendices D1 to D6, wherein the at least one processor performs a presentation control process for presenting historical information showing a history related to an insurance payment event, and a reception process for receiving input of an explanatory text describing the presented historical information, and in the extraction process, the at least one processor extracts related payment events related to the target event, treating the pair of the historical information and the explanatory text as a single target event.

[0173] (Note D8) An information processing device comprising at least one processor, wherein the at least one processor performs an acquisition process to acquire event information indicating a target event which is an event that may constitute an event that constitutes

[0174] (Note E1) A non-temporary recording medium that records a support program that causes a computer to perform an acquisition process to acquire event information indicating target events which are events that may constitute grounds for insurance payment, and an extraction process that uses a machine learning extraction model that takes a pair of documents and data as input and outputs the portion of the document in which the data is relevant, to extract related payment grounds related to the target events from a document that describes grounds for insurance payment.

[0175] (Note E2) A non-temporary recording medium that records a support program that causes a computer to execute an acquisition process for acquiring event information indicating target events which are events that may constitute an event that warrants payment of insurance benefits, and a determination process for determining whether or not the target events constitute an event that warrants payment using a language model that has been trained on natural language. [Explanation of Symbols]

[0176] 1. Information Processing Device 101 Acquisition unit (acquisition means) 102 Extraction part (extraction means) 1A Information Processing Device 101A Acquisition part (acquisition means) 102A Extraction part (extraction means) 103A Judgment unit (judgment means) 104A Presentation Control Unit (Presentation Control Means) 105A Reception area (reception method) M1 Extraction Model M2 Language Model

Claims

1. A means for obtaining event information that indicates a target event which is an event that may be a cause for payment of insurance benefits, An information processing device comprising: an extraction means for extracting relevant payment events related to the target event from a document describing the events for which insurance payments are made, using an extraction model trained on machine learning to take a pair of documents and data as input and output the portion of the document related to the data in the document.

2. A determination means for determining whether the aforementioned event falls under the aforementioned related payment event using a language model that has been trained on natural language, The information processing apparatus according to claim 1, further comprising a presentation control means for presenting the determination result of the determination means.

3. The information processing apparatus according to claim 2, wherein the determination means includes the related payment event and the target event, generates a prompt instructing the system to infer the relationship between the related payment event and the target event, and determines whether the target event corresponds to the related payment event based on the output obtained by inputting the generated prompt to the language model.

4. The determination means generates a prompt that instructs the system to output the basis for the inference, along with the inference result regarding the relationship between the relevant payment event and the target event. The information processing apparatus according to claim 3, wherein the presentation control means presents the inference result output by the language model or the determination result of the determination means together with the basis.

5. The system includes a receiving means for receiving instructions to modify the aforementioned inference results, The information processing apparatus according to claim 4, wherein the determination means re-determines whether the target event falls under the related payment event based on the correction instruction received by the receiving means.

6. The aforementioned receiving means receives correction instructions that express the content of the correction in natural language, The information processing apparatus according to claim 5, wherein the determination means includes a correction instruction received by the receiving means, generates a prompt instructing the determination means to infer the relationship between the target event and the related payment event based on the correction instruction, and re-determines whether the target event corresponds to the related payment event based on the output obtained by inputting the generated prompt to the language model.

7. A presentation control means that presents historical information showing the history related to the event of insurance payment, The system includes a receiving means for receiving input of an explanatory text describing the presented historical information, The information processing apparatus according to any one of claims 1 to 6, wherein the extraction means extracts related payment events associated with the target event, treating the set of historical information and the explanatory text as a single target event.

8. A means for obtaining event information that indicates a target event which is an event that may be a cause for payment of insurance benefits, An information processing device comprising: determination means for determining whether the aforementioned event constitutes a payment event using a language model that has been trained on natural language; and

9. At least one processor, A process to acquire event information that indicates a target event that may be an event that triggers the payment of insurance benefits, A support method for performing an extraction process that extracts relevant payment events related to the target event from a document describing the events for which insurance payments are made, using an extraction model trained on machine learning to take a pair of documents and data as input and output the portion of the document related to the data in that document.

10. Computers, A means for acquiring event information that indicates a target event which is an event that may constitute an event that triggers the payment of insurance benefits, and A support program that functions as an extraction means for extracting relevant payment events related to the target event from a document describing the events for which insurance payments are made, using an extraction model trained on machine learning to take a pair of documents and data as input and output the portion of the document in question that is related to the data.

Images