Claim settlement behavior prediction method and device, electronic equipment and storage medium
By automating the processing of claims application materials through blockchain networks and behavioral risk assessment models, the problem of insurance companies struggling to quickly verify the authenticity of claims documents has been solved, enabling rapid and accurate prediction of claims behavior and early warning of anomalies.
Patent Information
- Application Number
- CN202511624366.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-06
- Publication Date
- 2026-02-03
AI Technical Summary
Insurance companies struggle to quickly verify the authenticity of users' claim documents, making it difficult to detect potential abnormal claim behavior in a timely manner and affecting the efficiency of claim behavior prediction.
By using a blockchain network to assess the authenticity of information and combining it with a behavioral risk assessment model, claims application materials can be processed automatically, anomaly alerts can be generated, and human error and delays can be reduced.
It enables rapid and accurate assessment of the authenticity of claims requests, improves the accuracy of claims behavior prediction, reduces human error and delay risks, and enhances the ability to issue early warnings of abnormal claims behavior.
Smart Images
Figure CN121457722A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of artificial intelligence technology, and is applicable to the financial technology field, particularly to a method and device for predicting claims behavior, an electronic device, and a storage medium. Background Technology
[0002] In the financial insurance sector, when a user makes a claim, they need to provide relevant supporting documents. Subsequently, the insurance company's corresponding auditors need to verify these documents to determine whether the claim is genuine and credible, thereby avoiding the risk of being defrauded of premiums.
[0003] However, this approach presents challenges for insurance companies. Other supporting documents besides the policy are typically archived by different institutions and involve user privacy. Insurance companies cannot directly access these documents or must proactively contact the relevant institutions for verification. For example, in the process of claiming for health insurance, users provide their policies and medical records. However, because medical records are private information stored in hospital systems, insurance companies cannot quickly obtain and verify their authenticity, making it difficult to detect potential abnormal claims behavior in a timely manner. Therefore, improving the efficiency of predicting user claims behavior has become a pressing technical problem. Summary of the Invention
[0004] The main objective of this application is to provide a method, apparatus, electronic device, and storage medium for predicting claims behavior, aiming to improve the efficiency of predicting user claims behavior.
[0005] To achieve the above objectives, a first aspect of this application proposes a method for predicting claims behavior, the method comprising: Obtain claims requests and claim application materials from the target group; In response to the claim request, obtain the type of insurance for the claim; Based on the type of insurance claim, a target blockchain is selected from a preset blockchain network, and the authenticity of the claim application materials is evaluated based on the target blockchain to obtain a data authenticity score. Based on the target blockchain, a target behavior risk assessment model is selected from the preset original behavior risk assessment models; The target behavior risk assessment model is used to assess the behavior risk of the claim application materials and the claim request, and a claim behavior risk score is obtained. Based on the data authenticity score and the claims behavior risk score, the claims behavior of the target object is predicted to obtain the claims behavior type; wherein, the claims behavior type includes abnormal claims behavior or normal claims behavior.
[0006] In some embodiments, the claim application materials include document sub-data and material sub-data. The step of evaluating the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score includes: The first target node is determined from the blockchain network based on the single certificate data; The single-certificate sub-data is verified by the first target node and the target blockchain to obtain first verification information; wherein, the first verification information is used to indicate whether there is data identical to the single-certificate sub-data in the target blockchain; The second target node is determined from the blockchain network based on the material type of the material sub-data; The material sub-data is verified by comparing it with the second target node and the target blockchain to obtain second verification information; wherein, the second verification information is used to indicate whether there is data in the target blockchain that is the same as the material sub-data. The data authenticity score is obtained by calculating a score based on the first verification information and the second verification information.
[0007] In some embodiments, after calculating the data authenticity score based on the first verification information and the second verification information, the method further includes: If the claim behavior type is abnormal claim behavior, the risk level is assessed based on the data authenticity score and the claim behavior risk score to obtain the abnormal risk level. If the abnormal risk level is high risk, an early warning information is generated based on the data authenticity score, the claims behavior risk score, the document sub-data, and the material sub-data to obtain abnormal early warning information. Send the abnormal warning information to the first target node.
[0008] In some embodiments, the step of generating early warning information based on the data authenticity score, the claims behavior risk score, the document sub-data, and the material sub-data to obtain abnormal early warning information includes: Risk material data is selected from the document sub-data and the material sub-data based on the data authenticity score and the claims behavior risk score; If the risk material data includes the material sub-data, then the material sub-data is desensitized to obtain the target material sub-data; Anomaly warning information is generated based on the target material sub-data.
[0009] In some embodiments, the step of predicting the claim behavior of the target object based on the data authenticity score and the claim behavior risk score to obtain the claim behavior type includes: If at least one of the data authenticity score and the claims behavior risk score meets a preset anomaly judgment condition, the claims behavior type is determined to be an abnormal claims behavior; wherein, the anomaly judgment condition includes: the data authenticity score is greater than or equal to a preset authenticity score threshold, and the claims behavior risk score is greater than or equal to a preset risk score threshold. If neither the data authenticity score nor the claims behavior risk score meets the abnormal judgment criteria, the claims behavior type will be determined as normal claims behavior.
[0010] In some embodiments, after obtaining the claim request and claim application materials from the target object, the method further includes: Obtain the number of claims requests made by the target object within a preset time period; If the number of claim requests is greater than or equal to a preset request number threshold, the claim behavior type will be determined as abnormal claim behavior. If the number of claim requests is less than the request count threshold, then respond to the claim request and obtain the type of insurance for the claim.
[0011] In some embodiments, obtaining the claim request and claim application information from the target object includes: Obtain claims requests and original application materials from the target entity; Sensitive data identification is performed on the original application materials to obtain candidate sub-data; The candidate sub-data is encrypted to obtain the target sub-data; The original application information is updated based on the target sub-data to obtain the claim application information.
[0012] To achieve the above objectives, a second aspect of this application provides a claims behavior prediction device, the device comprising: The data acquisition module is used to acquire claim requests and claim application materials from the target object; The insurance type acquisition module is used to acquire the type of insurance for the claim in response to the claim request; The authenticity assessment module is used to select a target blockchain from a preset blockchain network based on the type of insurance claim, and to assess the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score. The risk assessment model screening module is used to screen out the target behavior risk assessment model from the preset original behavior risk assessment models based on the target blockchain; The risk assessment module is used to conduct a behavioral risk assessment on the claim application materials and the claim request through the target behavioral risk assessment model, and obtain a claim behavior risk score; The behavior prediction module is used to predict the claim behavior of the target object based on the data authenticity score and the claim behavior risk score, and obtain the claim behavior type; wherein, the claim behavior type includes abnormal claim behavior or normal claim behavior.
[0013] To achieve the above objectives, a third aspect of the present application provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method described in the first aspect.
[0014] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method described in the first aspect.
[0015] The claims behavior prediction method, apparatus, electronic device, and storage medium proposed in this application acquire the claims request and claims application materials of the target object, and in response to the claims request, determine the type of insurance for which the target object needs to make a claim. Then, it selects a target blockchain from a pre-set blockchain network to assess the authenticity of the data, and further filters out the corresponding behavioral risk assessment model based on the assessment score. Finally, it uses this model to assess the behavioral risk of the claims request and materials, obtaining a claims behavior risk score. Subsequently, it predicts the claims behavior of the target object, determining whether the claims process initiated by the target object constitutes abnormal claims behavior. The claims behavior prediction method provided in this application, combined with the decentralized characteristics of blockchain, can avoid information delays caused by cross-institutional trust issues, thereby enabling rapid and accurate assessment of the authenticity of claims requests and determining whether claims materials have been tampered with. Then, it uses a pre-trained target behavioral risk assessment model to further analyze the risks of the claims materials, improving the early warning capability for abnormal claims behavior during the claims process. This method can ultimately reduce the risk of human error or delay in review, effectively improve the accuracy of claims behavior prediction, and avoid potential risks to insurance companies due to their inability to quickly verify the authenticity of documents. Attached Figure Description
[0016] Figure 1 This is a flowchart of the claims behavior prediction method provided in the embodiments of this application; Figure 2 yes Figure 1 The flowchart of step S101 in the text; Figure 3This is another flowchart of the claims behavior prediction method provided in the embodiments of this application; Figure 4 yes Figure 1 The flowchart of step S103 in the process; Figure 5 This is another flowchart of the claims behavior prediction method provided in the embodiments of this application; Figure 6 yes Figure 5 The flowchart of step S502 in the document; Figure 7 This is a schematic diagram of the structure of the claims behavior prediction device provided in the embodiments of this application; Figure 8 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation
[0017] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0018] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.
[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.
[0020] First, let's analyze some of the terms used in this application: Artificial intelligence (AI) is a new branch of computer science that studies, develops, and applies theories, methods, technologies, and systems to simulate, extend, and expand human intelligence. It aims to understand the essence of intelligence and produce intelligent machines that can react in a way similar to human intelligence. Research in this field includes robotics, speech recognition, image recognition, natural language processing, and expert systems. AI can simulate the information processes of human consciousness and thought. Furthermore, AI utilizes digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceiving the environment, acquiring knowledge, and using that knowledge to achieve optimal results.
[0021] Natural Language Processing (NLP): NLP uses computers to process, understand, and utilize human language (such as Chinese and English). It is a branch of artificial intelligence and an interdisciplinary field of computer science and linguistics, often referred to as computational linguistics. NLP includes syntactic analysis, semantic analysis, and discourse understanding. It is commonly used in machine translation, handwritten and printed character recognition, speech recognition and text-to-speech conversion, intent recognition, information extraction and filtering, text classification and clustering, sentiment analysis, and opinion mining. It involves data mining, machine learning, knowledge acquisition, knowledge engineering, artificial intelligence research, and linguistic research related to language computation.
[0022] Information extraction is a text processing technique that extracts factual information such as entities, relationships, and events from natural language text and outputs it as structured data. Information extraction is a technique for extracting specific information from text data. Text data is composed of specific units, such as sentences, paragraphs, and chapters. Text information is composed of smaller, specific units, such as characters, words, phrases, sentences, paragraphs, or combinations of these units. Extracting noun phrases, names of people, and place names from text data is an example of text information extraction. Of course, text information extraction techniques can extract information of various types.
[0023] Blockchain: A decentralized distributed ledger that is block-based, immutable, secure, and reliable. It combines distributed storage, peer-to-peer transmission, consensus mechanisms, cryptography, and other technologies to record transactions and information through a continuously growing chain of data blocks, ensuring data security and transparency.
[0024] In the financial insurance sector, when a user makes a claim, they need to provide relevant supporting documents. Subsequently, the insurance company's corresponding auditors need to verify these documents to determine whether the claim is genuine and credible, thereby avoiding the risk of being defrauded of premiums.
[0025] However, this approach presents challenges for insurance companies. Other supporting documents besides the policy are typically archived by different institutions and involve user privacy. Insurance companies cannot directly access these documents or must proactively contact the relevant institutions for verification. For example, in the process of claiming for health insurance, users provide their policies and medical records. However, because medical records are private information stored in hospital systems, insurance companies cannot quickly obtain and verify their authenticity, making it difficult to detect potential abnormal claims behavior in a timely manner. Therefore, improving the efficiency of predicting user claims behavior has become a pressing technical problem.
[0026] Based on this, embodiments of this application provide a method and apparatus for predicting claims behavior, an electronic device and a storage medium, which aim to improve the efficiency of predicting user claims behavior.
[0027] The claims behavior prediction method, apparatus, electronic device, and storage medium provided in this application are specifically described through the following embodiments. First, the claims behavior prediction method in this application embodiment is described.
[0028] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.
[0029] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.
[0030] The claims behavior prediction method provided in this application relates to the field of artificial intelligence technology. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application implementing the claims behavior prediction method, but is not limited to the above forms.
[0031] This application can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.
[0032] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.
[0033] Figure 1 This is an optional flowchart of the claims behavior prediction method provided in the embodiments of this application. Figure 1 The method may include, but is not limited to, steps S101 to S106.
[0034] Step S101: Obtain the claim request and claim application materials from the target object.
[0035] Step S102: In response to the claim request, obtain the type of insurance for the claim.
[0036] Step S103: Select a target blockchain from the preset blockchain network based on the type of insurance claim, and evaluate the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score.
[0037] Step S104: Select the target behavior risk assessment model from the preset original behavior risk assessment models based on the target blockchain.
[0038] Step S105: The claim application materials and claim requests are assessed for behavioral risks using the target behavioral risk assessment model to obtain a claim behavior risk score.
[0039] Step S106: Based on the data authenticity score and the claims behavior risk score, predict the claims behavior of the target object to obtain the claims behavior type; wherein, the claims behavior type includes abnormal claims behavior or normal claims behavior.
[0040] Steps S101 to S106 of this embodiment involve obtaining the claim request and claim application materials of the target object, and in response to the claim request, determining the type of insurance for which the target object needs to claim. Then, a target blockchain is selected from a pre-set blockchain network for data authenticity assessment. Based on the assessment score, a corresponding behavioral risk assessment model is further selected. Finally, the claim request and materials are used to conduct a behavioral risk assessment, resulting in a claim behavior risk score. Subsequently, the claim behavior of the target object is predicted to determine whether the claim process initiated by the target object constitutes abnormal claim behavior. The claim behavior prediction method provided in this embodiment, combined with the decentralized characteristics of blockchain, avoids information delays caused by cross-institutional trust issues, thereby enabling rapid and accurate assessment of the authenticity of claim requests and determining whether claim materials have been tampered with. Subsequently, a pre-trained target behavioral risk assessment model is used to further analyze the risk of claim materials, improving the early warning capability for abnormal claim behavior during the claim process. This method can ultimately reduce the risk of human error or delay in review, effectively improve the accuracy of claims behavior prediction, and avoid potential risks to insurance companies due to their inability to quickly verify the authenticity of documents.
[0041] In step S101 of some embodiments, the target object refers to the user who submits a claim or their agent. A claim request refers to a request made by the user to the insurance company regarding a specific claim matter, typically including information such as the type of insurance for which the claim is being made, a description of the claim, an accident description, and the time of the accident. For example, the target object inputs the text "I want to apply for a health insurance claim and I hope the insurance company will pay me: xx yuan." Claim application materials refer to the relevant supporting documents provided by the user to support their claim request, typically including but not limited to policy information, medical records, invoices, and loss lists.
[0042] The methods for obtaining claim requests and supporting documentation include: the claimant submits their claim request to the insurance company's claims platform via text or voice input, and provides corresponding supporting documents. On the claims platform, the user fills in the necessary claim information and uploads the required documents; the platform system automatically records this information and generates a claim request.
[0043] Please see Figure 2 In some embodiments, the files submitted by the user may contain information that involves user privacy but is unrelated to the information required for the claims process, such as the user's home address, social information, and voice characteristics. Step S101 may include, but is not limited to, steps S201 to S204: Step S201: Obtain the claim request and original application materials from the target object.
[0044] Step S202: Sensitive data identification is performed on the original application materials to obtain candidate sub-data.
[0045] Step S203: Encrypt the candidate sub-data to obtain the target sub-data.
[0046] Step S204: Update the original application information based on the target sub-data to obtain the claim application information.
[0047] In step S201 of some embodiments, the specific embodiments regarding the claim request have been explained in detail in the specific embodiments of step S101, so they will not be repeated here.
[0048] Original application materials refer to all original supporting documents provided by the user to support their claim, which are not subject to data processing. These documents typically include, but are not limited to, medical records, accident reports, medical invoices, and vehicle repair invoices.
[0049] In step S202 of some embodiments, candidate sub-data refers to all parts containing sensitive information identified from the original application materials, and the sensitive information in these parts does not affect the verification of claim information. Candidate sub-data may be ID card number, bank card number, home address, etc.
[0050] If the original application materials are text data, natural language processing technology can be used to identify sensitive information in the original application materials that will not affect the verification of the claims process. If the original application materials are image data (such as images of medical records, images of car accident scenes, images reflecting property damage, etc.), sensitive data can be identified using a pre-trained image recognition model.
[0051] In step S203 of some embodiments, the target sub-data is sensitive data obtained by encrypting the candidate sub-data, ensuring that it is not leaked or tampered with during storage and transmission. The purpose of encryption is to ensure that sensitive data can only be accessed by authorized personnel during the review process, and to conduct claims review while protecting privacy.
[0052] Encryption methods may include, but are not limited to, using homomorphic encryption, quantum encryption, and encrypting candidate subdata using the SM7 cryptographic algorithm (Commercial Cryptographic Algorithm No. 7).
[0053] In step S204 of some embodiments, the sensitive parts of the original claim materials are replaced with encrypted sensitive information to generate new claim application materials. The updated materials contain encrypted sensitive information, allowing the insurance company to continue the claim review without exposing the original data. For example, in the health insurance claim process, the user's submitted medical records are encrypted, and the encrypted medical record information is then used to replace the original claim application materials to generate new claim application materials. During the review, the insurance company only sees the encrypted information and cannot access the original sensitive data, thus ensuring privacy and security.
[0054] In some embodiments, users may input claims requests via voice. The voice can be converted into text to obtain voice-text, which is then used as the final claims request to prevent the user's voice characteristics from being leaked.
[0055] Steps S201 to S204, as illustrated in this embodiment, ensure the security of user privacy information by identifying and encrypting sensitive data in the claim application materials. Automated sensitive data identification and encryption avoids the risk of information leakage that may occur during manual review, thus improving the security and accuracy of claim review.
[0056] Please see Figure 3 In some embodiments, after step S101, the claims behavior prediction method provided in this embodiment may also include, but is not limited to, steps S301 to S303: Step S301: Obtain the number of claims requests made by the target object within a preset time period.
[0057] Step S302: If the number of claim requests is greater than or equal to a preset request number threshold, the claim behavior type is determined as abnormal claim behavior.
[0058] Step S303: If the number of claim requests is less than the request count threshold, then respond to the claim request and obtain the type of insurance for the claim.
[0059] In step S301 of some embodiments, the preset time period refers to a time period within a specific time range (e.g., one day, one week, one month, etc.) for statistical analysis of the number of claims requests. The number of claims requests refers to the total number of claims requests submitted by the target object within this time period.
[0060] In step S302 of some embodiments, the request count threshold refers to the maximum reasonable number of claim requests set by the system within a preset time period. When the number of claim requests for a target object is greater than or equal to this threshold, the system determines it as abnormal claim behavior. This application does not strictly limit the specific value of the request count threshold. For example, the preset time period can be set to one day, and the request count threshold can be set to 10 times. If a target object submits 11 claim requests within one day, the user's last claim behavior is directly marked as abnormal claim behavior.
[0061] In step S303 of some embodiments, if the number of claims requests made by the target object within a preset time period does not reach the request number threshold, the normal claims process continues and step S102 is executed.
[0062] Steps S301 to S303 shown in this application embodiment, by statistically analyzing the number of claim requests within a specific time period, can promptly detect abnormal situations where users frequently submit claim requests, thereby improving the ability to identify potential fraudulent behavior.
[0063] In some embodiments, if the same claim application materials are identified as being submitted multiple times, the claim behavior type is directly identified as abnormal claim behavior.
[0064] In step S102 of some embodiments, the type of insurance for claims refers to the type of insurance associated with the claim request, such as motor vehicle insurance, personal accident insurance, and health insurance. The type of insurance for claims can be determined by analyzing the claim request using keyword matching technology or a pre-trained semantic recognition model.
[0065] In step S103 of some embodiments, the preset blockchain network refers to a blockchain network that is pre-built and stores information related to claims. Since different types of insurance require different participating institutions, the nodes participating in the blockchain also differ. For example, auto insurance claims may involve insurance companies, traffic management agencies, and vehicle repair shops, while critical illness insurance involves insurance companies and hospitals. The target blockchain refers to the blockchain corresponding to the type of insurance claim.
[0066] The data authenticity score is a rating obtained after assessing the authenticity of the claim application materials, reflecting the credibility of the materials. For the specific process of the authenticity assessment, please refer to [link / reference needed]. Figure 4 In some embodiments, the claim application materials include document sub-data and material sub-data, and step S103 may include, but is not limited to, steps S401 to S405: Step S401: Determine the first target node from the blockchain network based on the certificate data.
[0067] Step S402: Perform data verification on the document sub-data according to the first target node and the target blockchain to obtain the first verification information; wherein, the first verification information is used to indicate whether there is data in the target blockchain that is the same as the document sub-data.
[0068] Step S403: Determine the second target node from the blockchain network based on the material type of the material sub-data.
[0069] Step S404: Perform data verification on the material sub-data according to the second target node and the target blockchain to obtain second verification information; wherein, the second verification information is used to indicate whether there is data in the target blockchain that is the same as the material sub-data.
[0070] Step S405: Calculate the score based on the first verification information and the second verification information to obtain the data authenticity score.
[0071] In step S401 of some embodiments, the document sub-data refers to the policy data in the claim application materials, typically including information such as the policy number, insurance type, and insured amount. The material sub-data refers to other supporting documents or materials, such as medical records, medical expense receipts, and repair records. The first target node refers to the node in the blockchain network that provides and uploads the relevant information of the document sub-data, typically the insurance company or insurance agent responsible for the policy. Based on the policy information in the claim application, the insurance company or insurance agent related to the policy is automatically identified, thereby locating the corresponding blockchain node of the insurance company in the blockchain network.
[0072] In step S402 of some embodiments, the first verification information refers to data consistency information obtained by the first target node based on data in the target blockchain, regarding whether the document sub-data exists in the target blockchain. The first target node compares the document sub-data with the policy information in the blockchain network to determine whether the policy information is consistent with the record in the blockchain. If they are consistent, the first verification information is: information consistent with the policy information exists in the target blockchain, indicating that the authenticity of the document sub-data is confirmed, and it can be considered that the data verification of the policy sub-data has passed at the first target node. If they are inconsistent, the first verification information is: data verification fails, indicating that the user is suspected of tampering with the information.
[0073] In step S403 of some embodiments, the second target node refers to a node in the blockchain network that stores information related to the material sub-data, typically the organization or unit providing the relevant materials. The determination of the second target node is based on the material type of the material sub-data. For example, if the material type of the material sub-data is a medical record, the hospital that issued the medical record is determined based on the information in the medical record, and the node associated with that hospital in the blockchain network is determined as the second target node. If the material sub-data is a repair record issued by an auto repair shop, then the node associated with that auto repair shop in the blockchain network is determined as the second target node.
[0074] In step S404 of some embodiments, the second verification information refers to the data consistency information obtained by the second target node based on the data in the target blockchain, regarding whether the material sub-data exists in the target blockchain.
[0075] The second target node compares the material sub-data with relevant records in the blockchain network to determine if the material sub-data exists and is consistent with the blockchain record. If they are consistent, the second verification information is: information consistent with the material sub-data exists in the target blockchain, indicating that the authenticity of the material sub-data is confirmed, and it can be considered that the policy sub-data has passed the data verification at the first target node. If they are inconsistent, the second verification information is: data verification fails, indicating that the user is suspected of tampering with the information. For example, if the repair amount in the repair invoice information is inconsistent with the repair order data recorded in the target blockchain, the second verification information is invalid, indicating that the repair invoice data provided by the user may have been tampered with.
[0076] Understandably, having different institutions verify the data for different materials can prevent the leakage of user privacy.
[0077] In step S405 of some embodiments, if the verification information indicates that the data in the target blockchain is consistent with the claim application materials provided by the target object, a certain authenticity score can be determined. The corresponding authenticity score is determined based on the verification information output by all nodes, and all authenticity scores are summed; the sum is the data authenticity score. For example, if the verification information indicates that the data in the target blockchain is consistent with the claim application materials provided by the target object, the corresponding authenticity score is 20. If the verification information indicates that the data in the target blockchain is inconsistent with the claim application materials provided by the target object, the corresponding authenticity score is 0. The target blockchain has three nodes, including one first target node and two second target nodes. The first verification information and one second verification information indicate that the data verification passed, while the other second verification information indicates that the data verification failed; the corresponding data authenticity score is 40.
[0078] Steps S401 to S405 as illustrated in this application embodiment, through multi-node verification of the blockchain network, ensure the authenticity of claims data while avoiding cross-institutional leakage of user privacy, reducing the possibility of human intervention, and improving the transparency and fairness of the review process.
[0079] In step S104 of some embodiments, the original behavioral risk assessment model refers to the risk assessment model obtained by performing federated learning on block data of any blockchain in the blockchain network. That is, each blockchain corresponds to an independent original behavioral risk assessment model in the blockchain network. The target behavioral risk assessment model refers to the original behavioral risk assessment model corresponding to the target blockchain.
[0080] In step S105 of some embodiments, the claim application materials and claim request are input into the target behavioral risk assessment model, which outputs a claim behavior risk score. The behavioral risk assessment model is needed to assess the risk of the claim application materials because even if the data provided by the user is consistent with the data provided by the node, it does not rule out the possibility that the claim application materials are not falsified. For example, in a health insurance claim, a doctor may intentionally modify the diagnosis in the medical record to assist the user in obtaining compensation. By assessing the risk of the user's medical records or medical images through the target behavioral risk assessment model, it can identify discrepancies between the indicator data of multiple examination items in the medical record and the symptoms that should be reflected in the corresponding diagnosis. In this case, the behavioral risk assessment model will assign a higher risk score based on this inconsistency, thereby indicating that the claim application has a high risk.
[0081] In step S106 of some embodiments, abnormal claims behavior refers to situations where the claims application or materials provided by the user during the claims process clearly do not meet normal claims standards. This typically manifests as providing false or altered evidence, frequent claims, or claims requests that do not match the actual situation. Normal claims behavior, on the other hand, refers to situations where the user provides true and legal claims materials in accordance with the insurance contract, meeting the insurance claims standards. Normal claims behavior is usually supported by true and reasonable evidence, and the claims request is consistent with the actual accident or loss, complying with the requirements of the insurance terms.
[0082] The claim behavior prediction can be implemented as follows: if at least one of the data authenticity score and the claim behavior risk score meets a preset anomaly judgment condition, the claim behavior type is determined to be an abnormal claim behavior. The anomaly judgment conditions include: the data authenticity score being greater than or equal to a preset authenticity score threshold, and the claim behavior risk score being greater than or equal to a preset risk score threshold. If neither the data authenticity score nor the claim behavior risk score meets the anomaly judgment condition, the claim behavior type is determined to be a normal claim behavior.
[0083] Please see Figure 5 In some embodiments, after step S106, early warning information can be generated based on abnormal claim application materials. The claim behavior prediction method of this embodiment may also include, but is not limited to, steps S501 to S503: Step S501: If the claim behavior type is abnormal claim behavior, assess the risk level based on the data authenticity score and the claim behavior risk score to obtain the abnormal risk level.
[0084] Step S502: If the abnormal risk level is high risk, generate early warning information based on data authenticity score, claims behavior risk score, document sub-data and material sub-data to obtain abnormal early warning information.
[0085] Step S503: Send an anomaly warning message to the first target node.
[0086] In step S501 of some embodiments, the abnormal risk level may include low risk and high risk, and the number of abnormal risk levels can be adjusted according to actual needs. For example, if the data authenticity score is lower than a preset authenticity score threshold, and the claim behavior risk score is higher than a preset risk score threshold, then the risk level of the claim behavior will be rated as high risk.
[0087] In step S502 of some embodiments, the abnormal warning information typically includes a detailed description of the claims behavior, the abnormal material data, the potential risk points, and handling suggestions.
[0088] Specifically, please refer to Figure 6In some embodiments, step S502 includes, but is not limited to, steps S601 to S603: Step S601: Select risk material data from the document sub-data and material sub-data based on the data authenticity score and the claims behavior risk score.
[0089] Step S602: If the risk material data includes material sub-data, then the material sub-data is desensitized to obtain the target material sub-data.
[0090] Step S603: Generate anomaly warning information based on the target material sub-data.
[0091] In step S601 of some embodiments, risk material data refers to material data that is determined to have high risk or abnormal behavior. Specifically, preset authenticity score thresholds and risk score thresholds can be set. For document sub-data or material sub-data, if at least one of the following occurs: the data authenticity score is lower than the preset authenticity score threshold, and the claims behavior risk score is higher than the preset risk score threshold, then the corresponding sub-data is regarded as risk material data.
[0092] In step S602 of some embodiments, since the subsequent abnormal warning information is sent to the first target node, i.e. to the insurance company, the policy data has already undergone preliminary desensitization processing when the user inputs it, so even if the risk material data includes the policy data, no further processing is required.
[0093] Because the data in question involves other user privacy information, only the fields or information showing anomalies can be provided to the insurance company. Other verified privacy data can then undergo further anonymization. For example, if the data in question is an image from a routine blood test, and one test result is abnormal, contradicting the expected symptoms, while the values of the other results are normal, the image area containing the other results can be masked, and the processed image can be used as the target data in question.
[0094] In step S603 of some embodiments, the anomaly warning information is generated based on the analysis results of the target material sub-data to alert claims reviewers to potential anomalies. For example, if the repair invoice in the target material sub-data does not match the historical records stored in the target blockchain, or if the same user submits multiple claims with duplicate repair items, an anomaly warning information is subsequently generated to prompt claims reviewers to further verify the authenticity of the claim request.
[0095] Steps S601 to S603, as illustrated in this embodiment, combine data authenticity scores and claims behavior risk scores to accurately screen high-risk material data and protect user privacy through anonymization. Simultaneously, the anomaly warning information generated based on candidate sub-data provides effective support for claims review, helping insurance companies to promptly identify potential fraudulent activities.
[0096] In step S503 of some embodiments, the abnormal warning information is sent to the first target node via a preset communication protocol. This information can be transmitted through smart contracts of the blockchain network, notification systems, or other suitable transmission methods.
[0097] Steps S501 to S503 as shown in the embodiments of this application reduce the cost and error rate of manual review by comprehensively evaluating the data authenticity score, the claims behavior risk score and the material data, thereby improving the efficiency and accuracy of claims review and enhancing the insurance company's ability to deal with potential fraud.
[0098] Please see Figure 7 This application also provides a claims behavior prediction device, which can implement the above-mentioned claims behavior prediction method. The device includes: Data acquisition module 701 is used to acquire claim requests and claim application materials from the target object; The insurance type acquisition module 702 is used to obtain the types of insurance for claims in response to claims requests; The authenticity assessment module 703 is used to select a target blockchain from a preset blockchain network based on the type of insurance claim, and to assess the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score. The risk assessment model screening module 704 is used to screen out the target behavior risk assessment model from the preset original behavior risk assessment models based on the target blockchain. Risk assessment module 705 is used to conduct behavioral risk assessment on claim application materials and claim requests through the target behavioral risk assessment model to obtain a claim behavior risk score; The behavior prediction module 706 is used to predict the claim behavior of the target object based on the data authenticity score and the claim behavior risk score, so as to obtain the claim behavior type; among which, the claim behavior type includes abnormal claim behavior or normal claim behavior.
[0099] The specific implementation of this claims behavior prediction device is basically the same as the specific implementation of the claims behavior prediction method described above, and will not be repeated here.
[0100] This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the aforementioned claims behavior prediction method. This electronic device can be any smart terminal, including tablet computers, in-vehicle computers, etc.
[0101] Please see Figure 8 , Figure 8 The hardware structure of an electronic device according to another embodiment is illustrated. The electronic device includes: The processor 801 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application. The memory 802 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 802 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 802 and is called and executed by the processor 801 using the claims behavior prediction method of the embodiments of this application. The 803 input / output interface is used to implement information input and output. The communication interface 804 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.). Bus 805 transmits information between various components of the device (e.g., processor 801, memory 802, input / output interface 803, and communication interface 804); The processor 801, memory 802, input / output interface 803, and communication interface 804 are connected to each other within the device via bus 805.
[0102] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the aforementioned claims behavior prediction method.
[0103] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.
[0104] The claims behavior prediction method, device, electronic device, and storage medium provided in this application obtain the claims request and application materials of the target object, and in response to the claims request, determine the type of insurance for which the target object needs to make a claim. Then, a target blockchain is selected from a preset blockchain network for data authenticity assessment. Based on the assessment score, a corresponding behavioral risk assessment model is further selected. Finally, the model is used to assess the behavioral risk of the claims request and materials, obtaining a claims behavior risk score. Subsequently, the claims behavior of the target object is predicted to determine whether the claims process initiated by the target object constitutes abnormal claims behavior. The claims behavior prediction method provided in this application combines the decentralized characteristics of blockchain, avoiding information delays caused by cross-institutional trust issues, thereby enabling rapid and accurate assessment of the authenticity of claims requests and determining whether claims materials have been tampered with. Subsequently, a pre-trained target behavioral risk assessment model is used to further analyze the risks of the claims materials, improving the early warning capability for abnormal claims behavior during the claims process. This method can ultimately reduce the risk of human error or delay in review, effectively improve the accuracy of claims behavior prediction, and avoid potential risks to insurance companies due to their inability to quickly verify the authenticity of documents.
[0105] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.
[0106] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.
[0107] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.
[0108] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.
[0109] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0110] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.
[0111] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.
[0112] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.
[0113] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.
[0114] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0115] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.
Claims
1. A method for predicting claims behavior, characterized in that, The method includes: Obtain claims requests and claim application materials from the target group; In response to the claim request, obtain the type of insurance for the claim; Based on the type of insurance claim, a target blockchain is selected from a preset blockchain network, and the authenticity of the claim application materials is evaluated based on the target blockchain to obtain a data authenticity score. Based on the target blockchain, a target behavior risk assessment model is selected from the preset original behavior risk assessment models; The target behavior risk assessment model is used to assess the behavior risk of the claim application materials and the claim request, and a claim behavior risk score is obtained. Based on the data authenticity score and the claims behavior risk score, the claims behavior of the target object is predicted to obtain the claims behavior type; wherein, the claims behavior type includes abnormal claims behavior or normal claims behavior.
2. The method according to claim 1, characterized in that, The claim application materials include document sub-data and material sub-data. The step of evaluating the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score includes: The first target node is determined from the blockchain network based on the single certificate data; The single-certificate sub-data is verified by the first target node and the target blockchain to obtain first verification information; wherein, the first verification information is used to indicate whether there is data identical to the single-certificate sub-data in the target blockchain; The second target node is determined from the blockchain network based on the material type of the material sub-data; The material sub-data is verified by comparing it with the second target node and the target blockchain to obtain second verification information; wherein, the second verification information is used to indicate whether there is data in the target blockchain that is the same as the material sub-data. The data authenticity score is obtained by calculating a score based on the first verification information and the second verification information.
3. The method according to claim 2, characterized in that, After calculating the data authenticity score based on the first verification information and the second verification information, the method further includes: If the claim behavior type is abnormal claim behavior, the risk level is assessed based on the data authenticity score and the claim behavior risk score to obtain the abnormal risk level. If the abnormal risk level is high risk, an early warning information is generated based on the data authenticity score, the claims behavior risk score, the document sub-data, and the material sub-data to obtain abnormal early warning information. Send the abnormal warning information to the first target node.
4. The method according to claim 3, characterized in that, The method of generating early warning information based on the data authenticity score, the claims behavior risk score, the document sub-data, and the material sub-data yields abnormal early warning information, including: Risk material data is selected from the document sub-data and the material sub-data based on the data authenticity score and the claims behavior risk score; If the risk material data includes the material sub-data, then the material sub-data is desensitized to obtain the target material sub-data; Anomaly warning information is generated based on the target material sub-data.
5. The method according to any one of claims 1 to 4, characterized in that, The method of predicting the claim behavior of the target object based on the data authenticity score and the claim behavior risk score to obtain the claim behavior type includes: If at least one of the data authenticity score and the claims behavior risk score meets a preset anomaly judgment condition, the claims behavior type is determined to be an abnormal claims behavior; wherein, the anomaly judgment condition includes: the data authenticity score is greater than or equal to a preset authenticity score threshold, and the claims behavior risk score is greater than or equal to a preset risk score threshold. If neither the data authenticity score nor the claims behavior risk score meets the abnormal judgment criteria, the claims behavior type will be determined as normal claims behavior.
6. The method according to any one of claims 1 to 4, characterized in that, After obtaining the claim request and claim application materials from the target object, the method further includes: Obtain the number of claims requests made by the target object within a preset time period; If the number of claim requests is greater than or equal to a preset request number threshold, the claim behavior type will be determined as abnormal claim behavior. If the number of claim requests is less than the request count threshold, then respond to the claim request and obtain the type of insurance for the claim.
7. The method according to any one of claims 1 to 4, characterized in that, The acquisition of claim requests and claim application materials from the target object includes: Obtain claims requests and original application materials from the target entity; Sensitive data identification is performed on the original application materials to obtain candidate sub-data; The candidate sub-data is encrypted to obtain the target sub-data; The original application information is updated based on the target sub-data to obtain the claim application information.
8. A claims behavior prediction device, characterized in that, The device includes: The data acquisition module is used to acquire claim requests and claim application materials from the target object; The insurance type acquisition module is used to acquire the type of insurance for the claim in response to the claim request; The authenticity assessment module is used to select a target blockchain from a preset blockchain network based on the type of insurance claim, and to assess the authenticity of the claim application materials based on the target blockchain to obtain a data authenticity score. The risk assessment model screening module is used to screen out the target behavior risk assessment model from the preset original behavior risk assessment models based on the target blockchain; The risk assessment module is used to conduct a behavioral risk assessment on the claim application materials and the claim request through the target behavioral risk assessment model, and obtain a claim behavior risk score; The behavior prediction module is used to predict the claim behavior of the target object based on the data authenticity score and the claim behavior risk score, so as to obtain the claim behavior type; wherein, the claim behavior type includes abnormal claim behavior or normal claim behavior.
9. An electronic device, characterized in that, The electronic device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the method according to any one of claims 1 to 7.
10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.