A blockchain-based information processing method and system

Through blockchain technology and machine learning models, the problems of inefficient efficiency, insufficient fraud detection and untimely fluctuation monitoring in insurance claims management are solved, automated processing and real-time monitoring are realized, and the efficiency of claim processing and anti-fraud capabilities are improved.

CN117541171BActive Publication Date: 2024-11-26SHANGHAI XINCUO TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311376905.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-10-23
Publication Date
2024-11-26
Estimated Expiration
2043-10-23

AI Technical Summary

Technical Problem

In the insurance claim management process, there are problems such as inefficient processing efficiency, insufficient fraud detection and untimely monitoring of claim amount fluctuations, resulting in waste of resources and potential risks.

Method used

Adopt blockchain-based information processing methods, by establishing an insurance claim blockchain network, collecting and analyzing claim data, using machine learning models to calculate behavior index, event contradiction index and abnormal fluctuation index, automated data processing and fraud detection, and real-time monitoring And issue a warning.

Benefits of technology

It improves the efficiency and accuracy of claim handling, reduces manual intervention, enhances anti-fraud capabilities, promptly identify and respond to abnormal situations, and optimizes resource allocation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117541171B_ABST
    Figure CN117541171B_ABST
Patent Text Reader

Abstract

The present invention discloses a blockchain-based information processing method and system, which relates to the technical field of insurance claim information processing. The system enables claims to be processed and paid more quickly through the application of automated data processing, feature extraction, and machine learning models. By analyzing the claimant's historical data and current claim information, abnormal behavior, contradictory information, and fraud risks can be more accurately identified. This helps to reduce the incidence of insurance fraud and protect the economic interests of insurance companies. Claim data is monitored in real time. Once an abnormality is found, the system can immediately issue an alert and take necessary actions to reduce potential risks. This helps insurance companies better manage risks and improves the effectiveness of risk management. The system optimizes resource allocation. Automatically marking claims that require additional review means that review resources can be more specifically allocated to cases that really need attention, avoiding waste of resources.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of insurance claim information processing, and specifically to an information processing method and system based on blockchain. Background Art

[0002] The insurance industry has always been an important part of the financial sector, providing risk protection and claims services. However, insurance claims management has always been a complex and cumbersome process that requires a lot of manpower and time to process claims information. Traditional insurance claims processing methods usually face the following problems:

[0003] (1) Traditional claims information processing usually involves a lot of documentation and manual data entry, which is prone to errors and requires a lot of time and resources;

[0004] (2) Some claims may involve fraud, where the claimant may exaggerate the loss or provide false information to obtain more compensation;

[0005] (3) Fluctuations in claim amounts may cause concern, but traditional methods are usually unable to identify abnormal fluctuations in a timely manner. To address these issues, traditional claims management methods have many shortcomings in processing efficiency, fraud detection, and fluctuation monitoring. To address these issues, new technologies and methods have become particularly important. It is urgent to propose a blockchain-based information processing method and system that may provide a potential solution to improve the efficiency, accuracy, and security of claims processing. Summary of the invention

[0006] 1. Technical issues to be resolved

[0007] In view of the shortcomings of the prior art, the present invention provides a blockchain-based information processing method and system to address the many shortcomings in processing efficiency, fraud detection and fluctuation monitoring proposed by the background technology.

[0008] (II) Technical solution

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions: a blockchain-based information processing method and system, comprising the following steps:

[0010] S1. Establish an insurance claim blockchain network, set up a blockchain consensus mechanism, make several nodes in the blockchain have a copy of the complete blockchain, and establish claim rules, conditions and payment contracts; and use it to verify and analyze claim data;

[0011] S2. Collect historical claim record information submitted by the claimant and the claimant’s current claim information;

[0012] S3. Extract features from historical claim record information, analyze claim frequency, historical claim amount, and whether there are contradictions in event descriptions, and use time series analysis methods to analyze changes in claim amounts over time and whether there are abnormal fluctuations. Establish a claim model, train a machine learning model based on historical fraud cases and normal claim cases, and analyze and calculate the current claim information of the claimant to obtain a behavior index Xw, an event contradiction index Md, and an abnormal fluctuation index Yc. The behavior index Xw, event contradiction index Md, and abnormal fluctuation index Yc are obtained by the following formula:

[0013]

[0014] In the formula, F represents the frequency value of the number of claims made by the claimant in the past year, A represents the total amount of claims made by the claimant in the past, Nt represents the number of different insurance types or risk types, E represents the expected value or average value of the insurance amount variable based on historical data, which is used to compare with the current claim amount. If the current claim amount is higher than the expected value, the claim event needs to be reviewed abnormally; Cd represents the number of contradictory words or contradictory information in the claim document, Td represents the total number of words in the claim document; At represents the floating value of the claim amount changing with the event, R represents the expected normal fluctuation range calculated based on historical data; C1 represents the correction constant;

[0015] S4, comparing the claimant's behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result;

[0016] S5. Review the claimant’s current claim information based on the forecast assessment results.

[0017] Preferably, the S1 specifically includes: establishing a blockchain platform of Ethereum, Hyperledger, Fabric or Corda, setting a data structure and storage method of claim data, and configuring a blockchain node, storing a copy of the complete blockchain in the blockchain node;

[0018] And choose Proof of Work, Proof of Stake or Byzantine Fault Tolerant consensus mechanism to ensure the consistency of claim data between nodes;

[0019] According to the needs of insurance claims, write smart contracts, perform functional tests, performance tests, and security tests on the smart contracts, and deploy the smart contracts to the blockchain nodes to ensure that the smart contracts are consistent with the claim rules and conditions.

[0020] Preferably, the historical claim record information includes claimant personal information, historical claim type information, historical claim amount information, event date and description information, claim processing result information and payment record information;

[0021] The claimant's current claim information includes event description information, current claim amount information, claim event date and additional information, current claim document information and claim status information. The claim document information includes medical certificates, accident reports and event photos.

[0022] Preferably, a claim model is established, and feature extraction and analysis are performed based on the claimant's historical claim record information, and the claim frequency F is calculated. The number of claims submitted for events in the past year is calculated using a time window method, and the starting point of the time window is determined. The window is sliding according to the selected time window size, and the number of claims submitted by the claimant during the window period is counted using a counting operation, and updated once a month;

[0023] Calculate the comprehensive, average, and standard deviation of the claimant's historical claim amounts to statistically analyze relevant indicators and obtain the distribution of the claimant's claim amounts;

[0024] Use natural language processing (NLP) technology, including text mining and text analysis, to analyze the language patterns of historical claim time descriptions for contradictions or anomalies, and perform statistics;

[0025] Based on the time series analysis method, moving average or exponential smoothing method is used to analyze the trend of claim amount over time; detect whether there are abnormal fluctuations and the value of abnormal fluctuations.

[0026] Preferably, the S4 specifically includes:

[0027] The behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant are respectively compared with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 to obtain the following prediction evaluation results;

[0028] The predicted evaluation results include normal results, event-contradictory results, abnormal fluctuation results, and multiple problem results;

[0029] The normal results include the following:

[0030] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0031] If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event;

[0032] If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal;

[0033] When all indices are below the threshold, the claim is considered normal and no additional processing is required.

[0034] Preferably, the conflicting results of the event include:

[0035] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0036] If the event contradiction index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required;

[0037] If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal;

[0038] Because the contradiction index Md exceeds the event safety threshold Q2, the claim requires additional review to resolve the contradiction.

[0039] Preferably, the abnormal fluctuation results include:

[0040] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0041] If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event;

[0042] If the abnormal fluctuation index Yc ≥ abnormal safety threshold Q3, it means that the claim amount fluctuates abnormally;

[0043] Because the contradiction index Md exceeds the event safety threshold Q2, the claim requires additional review to resolve the contradiction.

[0044] Because the abnormal volatility index Yc exceeds the abnormal safety threshold Q3, the claim requires additional review to investigate abnormal volatility.

[0045] Preferably, the multiple question results include:

[0046] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0047] If the abnormal fluctuation index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required;

[0048] Abnormal fluctuation index ≥ abnormal safety threshold Q3, indicating abnormal fluctuation of claim amount;

[0049] Since both the contradiction index Md and the abnormal fluctuation index Yc exceed the threshold, the claim requires detailed review to resolve the contradiction and abnormal fluctuation issues.

[0050] Preferably, according to the prediction evaluation results, the claimant's current claim information is considered normal, without abnormal behavior or contradiction, then the claim is processed and approved for payment; the review results show that the claimant's historical behavior is consistent with the current claim information, and no warning information needs to be triggered;

[0051] If the predictive assessment results indicate potential abnormal behavior, inconsistencies, or risks, but it is not yet certain whether the claim is normal, the claim information will be marked as requiring additional review and marked as a fraud alert, triggering a warning message.

[0052] A blockchain-based information processing system, comprising a blockchain network module, a data collection module, a feature extraction module, an analysis module, a behavior evaluation module, and a review module;

[0053] The blockchain network module is used to establish an insurance claim blockchain network, configure a blockchain consensus mechanism, ensure that nodes in the blockchain have a copy of the complete blockchain; and deploy smart contracts to manage the execution of claim data and rules;

[0054] The data collection module is used to collect historical claim record information and current claim information submitted by the claimant, and adopts a data interface mode and a front-end interface interaction mode for the claimant to submit information;

[0055] The feature extraction module is used to extract features from historical claim record information, including claim frequency, historical claim amount, and whether there is any contradiction in event description;

[0056] The analysis module is used to use time series analysis methods to analyze the changes in claim amounts over time, detect abnormal fluctuations, and establish claim models, including machine learning models, for fraud detection and analysis;

[0057] The behavior evaluation module is used to calculate the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant, and compare the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result;

[0058] The review module is used to review the claimant's current claim information based on the prediction and evaluation results, which include normal results, event-contradictory results, abnormal fluctuation results, and multiple-problem results.

[0059] (III) Beneficial effects

[0060] The present invention provides a blockchain-based information processing method and system, which has the following beneficial effects:

[0061] (1) Automates many tedious tasks and information, including data collection, feature extraction, behavior assessment, and anomaly detection. This reduces manual intervention and processing time, thereby improving the efficiency of the entire claims processing process and reducing the cost of processing claims.

[0062] (2) Through machine learning models and data analysis, the system can more accurately identify abnormal behavior, fraud risks, and abnormal fluctuations in claim amounts. This helps insurance companies review claims more accurately, reduce the occurrence of fraudulent claims, and improve anti-fraud capabilities.

[0063] (3) The system can monitor claims data in real time and immediately issue warnings and take appropriate measures if an anomaly is detected. This helps reduce potential risks and respond to problems in a timely manner.

[0064] (4) The system can automatically mark claims that require additional review, which means that review resources can be more specifically allocated to cases that really need attention, avoiding wasting resources on normal claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 This is a schematic diagram of a block diagram of an information processing system based on blockchain in the present invention. DETAILED DESCRIPTION

[0066] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0067] Traditional insurance claims processing methods usually face the following problems:

[0068] (1) Traditional claims information processing usually involves a lot of documentation and manual data entry, which is prone to errors and requires a lot of time and resources;

[0069] (2) Some claims may involve fraud, where the claimant may exaggerate the loss or provide false information to obtain more compensation;

[0070] (3) Fluctuations in claim amounts may cause concern, but traditional methods are usually unable to identify abnormal fluctuations in a timely manner. To address these issues, traditional claims management methods have many shortcomings in processing efficiency, fraud detection, and fluctuation monitoring. To address these issues, new technologies and methods have become particularly important. It is urgent to propose a blockchain-based information processing method and system that may provide a potential solution to improve the efficiency, accuracy, and security of claims processing.

[0071] Example 1

[0072] The present invention provides a blockchain-based information processing method and system, comprising the following steps:

[0073] S1. Establish an insurance claim blockchain network, set up a blockchain consensus mechanism, make several nodes in the blockchain have a copy of the complete blockchain, and establish claim rules, conditions and payment contracts; and use it to verify and analyze claim data;

[0074] S2. Collecting historical claim record information submitted by the claimant and the claimant’s current claim information;

[0075] S3. Extract features from historical claim record information, analyze claim frequency, historical claim amount, and whether there are contradictions in event descriptions, and use time series analysis methods to analyze changes in claim amounts over time and whether there are abnormal fluctuations. Establish a claim model, train a machine learning model based on historical fraud cases and normal claim cases, and analyze and calculate the current claim information of the claimant to obtain a behavior index Xw, an event contradiction index Md, and an abnormal fluctuation index Yc. The behavior index Xw, event contradiction index Md, and abnormal fluctuation index Yc are obtained by the following formula:

[0076]

[0077] In the formula, F represents the frequency value of the number of claims made by the claimant in the past year, A represents the total amount of claims made by the claimant in the past, Nt represents the number of different insurance types or risk types, E represents the expected value or average value of the insurance amount variable based on historical data, which is used to compare with the current claim amount. If the current claim amount is higher than the expected value, the claim event needs to be reviewed abnormally; Cd represents the number of contradictory words or contradictory information in the claim document, Td represents the total number of words in the claim document; At represents the floating value of the claim amount changing with the event, R represents the expected normal fluctuation range calculated based on historical data; C1 represents the correction constant;

[0078] By calculating the behavior index Xw, the event contradiction index Md, and the abnormal fluctuation index Yc, the system can automatically generate a risk assessment report for each claimant. This helps insurance companies quickly identify high-risk claim cases so that they can take necessary actions faster. Multiple parameters are used in the formula, including the number of historical claims, the total amount, document contradiction information, fluctuations in the claim amount, etc., which take into account the claimant's historical behavior and the characteristics of the current claim. This comprehensive analysis enables the system to detect abnormal behavior and fraud attempts more accurately. Automatic calculation of the behavior index Xw, the event contradiction index Md, and the abnormal fluctuation index Yc can significantly improve the efficiency of claims processing. The system can automatically generate an index for each claim, thereby reducing the time and cost required for manual intervention.

[0079] S4, comparing the claimant's behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result;

[0080] S5. Review the claimant’s current claim information based on the forecast assessment results.

[0081] In this embodiment, a lot of manual work and document processing are reduced through automation and smart contract execution. This helps to shorten the time of claim processing, improve efficiency, and reduce the waste of manpower and resources. The method uses machine learning models and behavioral index Xw calculation to more accurately identify potential fraudulent behavior. By analyzing the claimant's history and behavior patterns, abnormal situations can be detected in time, thereby reducing the risk of fraudulent claims. Using time series analysis methods to monitor fluctuations in claim amounts, abnormal fluctuations can be detected in time and automatically trigger additional reviews. This helps insurance companies better manage risks and reduce unnecessary losses.

[0082] Example 2

[0083] This embodiment is an explanation of the embodiment 1. Specifically, S1 includes: establishing a blockchain platform of Ethereum, Hyperledger, Fabric or Corda, setting the data structure and storage method of the claim data, and configuring the blockchain node, storing a copy of the complete blockchain in the blockchain node; these platforms have been widely used in enterprise-level applications, have scalability, stability and security, and help ensure the security and consistency of the claim data.

[0084] And choose Proof of Work, Proof of Stake or Byzantine Fault Tolerant consensus mechanism to ensure the consistency of claim data between nodes; this helps to ensure data consistency and easy retrieval. The distributed ledger structure of blockchain can provide a high degree of data traceability, which helps to monitor the entire life cycle of the claim.

[0085] According to the needs of insurance claims, smart contracts are written, and functional, performance and security tests are performed on the smart contracts. The smart contracts are then deployed to the blockchain nodes to ensure that the smart contracts are consistent with the claim rules and conditions. Functional, performance and security tests help to discover and correct potential problems and ensure the stability and reliability of smart contracts. Blockchain technology ensures the consistency and traceability of claim data. Each node stores a complete copy of the blockchain, and any data changes require consensus and are not easily tampered with. This helps eliminate disputes and errors in the claims process. Blockchain technology can automate the claims process, reducing the need for manual intervention and thus reducing operating costs. The automatic execution of smart contracts ensures the consistent application of claim rules and eliminates human errors.

[0086] In this embodiment, the above measures help to establish an efficient, secure, consistent and traceable insurance claims management system, which will improve the operational efficiency of the insurance business, reduce fraud risks and improve data security.

[0087] Example 3

[0088] This embodiment is an explanation of the embodiment 1. Specifically, the historical claim record information includes claimant personal information, historical claim type information, historical claim amount information, event date event and description information, claim processing result information and payment record information;

[0089] The claimant's current claim information includes event description information, current claim amount information, claim event date and additional information, current claim document information and claim status information. The claim document information includes medical certificates, accident reports and event photos.

[0090] Example 4

[0091] This embodiment is an explanation of the embodiment 1. Specifically, a claim model is established. According to the claimant's historical claim record information, feature extraction and analysis are performed to calculate the claim frequency F. The number of claims submitted for events in the past year is calculated using the time window method. The starting point of the time window is determined. The window is slid according to the selected time window size. The number of claims submitted by the claimant during the window period is counted using a counting operation and updated once a month.

[0092] Calculate the comprehensive, average, and standard deviation of the claimant's historical claim amounts to statistically analyze relevant indicators and obtain the distribution of the claimant's claim amounts;

[0093] Use natural language processing (NLP) technology, including text mining and text analysis, to analyze the language patterns of historical claim time descriptions for contradictions or anomalies, and perform statistics;

[0094] Based on the time series analysis method, moving average or exponential smoothing method is used to analyze the trend of claim amount over time; detect whether there are abnormal fluctuations and the value of abnormal fluctuations.

[0095] In this embodiment, the automated feature extraction and analysis process can greatly reduce the workload of manual processing. The monthly updated claim frequency and other indicators can be automatically calculated, reducing operating costs and improving efficiency. By comprehensively analyzing historical claim data, the system can more easily detect possible fraud, thereby reducing the risk of fraud and saving the insurance company's losses.

[0096] Example 5

[0097] This embodiment is explained in Embodiment 1. Specifically, S4 includes:

[0098] The behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant are respectively compared with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 to obtain the following prediction evaluation results;

[0099] The prediction and evaluation results include normal results, event contradiction results, abnormal fluctuation results and multiple problem results; by comparing different indexes with corresponding thresholds, the system can automatically classify claim cases into normal, event contradiction, abnormal fluctuation or multiple problems. This automatic classification helps insurance companies handle different types of claims more efficiently.

[0100] The normal results include the following:

[0101] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0102] If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event;

[0103] If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal;

[0104] When all indices are below the threshold, the claim is considered normal and no additional processing is required.

[0105] Specifically, the conflicting results of the event include:

[0106] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0107] If the event contradiction index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required;

[0108] If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal;

[0109] Because the contradiction index Md exceeds the event safety threshold Q2, the claim requires additional review to resolve the contradiction.

[0110] The abnormal fluctuation results include:

[0111] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0112] If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event;

[0113] If the abnormal fluctuation index Yc ≥ abnormal safety threshold Q3, it means that the claim amount fluctuates abnormally;

[0114] Because the contradiction index Md exceeds the event safety threshold Q2, the claim requires additional review to resolve the contradiction.

[0115] Because the abnormal volatility index Yc exceeds the abnormal safety threshold Q3, the claim requires additional review to investigate abnormal volatility.

[0116] The multiple question results include:

[0117] If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal;

[0118] If the abnormal fluctuation index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required;

[0119] Abnormal fluctuation index ≥ abnormal safety threshold Q3, indicating abnormal fluctuation of claim amount;

[0120] Since both the contradiction index Md and the abnormal fluctuation index Yc exceed the threshold, the claim requires detailed review to resolve the contradiction and abnormal fluctuation issues.

[0121] In this embodiment, by comparing the claimant's behavior index Xw, event contradiction index Md, and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2, and abnormal safety threshold Q3, the system can automatically classify the claim case into normal, event contradiction, abnormal fluctuation, or multiple problems. This automatic classification helps insurance companies to handle different types of claims more efficiently. By classifying fraud cases into event contradictions or multiple problems, the system helps insurance companies identify potential fraud earlier and take measures to reduce fraud losses.

[0122] Example 6

[0123] This embodiment is an explanation of the embodiment 1. Specifically, according to the prediction and evaluation results, the claimant's current claim information is considered normal, without abnormal behavior or contradiction, so the claim is processed and approved for payment; the claim information normal result indicates that the claimant's historical behavior is consistent with the current claim information, and no warning information needs to be triggered;

[0124] If the predictive assessment results indicate potential abnormal behavior, inconsistencies, or risks, but it is not yet certain whether the claim is normal, the claim information will be marked as requiring additional review and marked as a fraud alert, triggering a warning message.

[0125] Based on the predictive assessment results, the system can automatically flag claims that require additional review and provide fraud alerts. This reduces the need for manual intervention and ensures timely and accurate reviews.

[0126] Example 7

[0127] A blockchain-based information processing system, see Figure 1 , including blockchain network module, data collection module, feature extraction module, analysis module, behavior assessment module and review module;

[0128] The blockchain network module is used to establish an insurance claim blockchain network, configure a blockchain consensus mechanism, ensure that nodes in the blockchain have a copy of the complete blockchain; and deploy smart contracts to manage the execution of claim data and rules;

[0129] The data collection module is used to collect historical claim record information and current claim information submitted by the claimant, and adopts a data interface mode and a front-end interface interaction mode for the claimant to submit information;

[0130] The feature extraction module is used to extract features from historical claim record information, including claim frequency, historical claim amount, and whether there is any contradiction in event description;

[0131] The analysis module is used to use time series analysis methods to analyze the changes in claim amounts over time, detect abnormal fluctuations, and establish claim models, including machine learning models, for fraud detection and analysis;

[0132] The behavior evaluation module is used to calculate the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant, and compare the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result;

[0133] The review module is used to review the claimant's current claim information based on the prediction and evaluation results, which include normal results, event-contradictory results, abnormal fluctuation results, and multiple-problem results.

[0134] Specific data example 1:

[0135] F = 10 (the frequency value of the number of claims in the past year);

[0136] A = 5000 (total amount of past claims);

[0137] Nt = 5 (the number of different insurance types or risk types);

[0138] E = 1000 (the expected value or average value of the insurance amount variable based on historical data);

[0139] Cd = 2 (the number of contradictory words or contradictory information in the claim document);

[0140] Td = 20 (total number of words in the claim document);

[0141] At = 300 (the floating value of the claim amount as it changes with the event);

[0142] R = 50 (the expected normal fluctuation range calculated based on historical data);

[0143] C1=0.1 (corrected constant);

[0144] Substituting the above values ​​into the formula:

[0145] Behavior index Xw = (10 + 5000) / (5 + 1000) + 0.1 = 5.1;

[0146] Event contradiction index Md = 2 / 20 = 0.1;

[0147] Abnormal fluctuation index Yc = (300-50) / 50 = 5.0;

[0148] In this example, the behavior index Xw is 5.1, the event contradiction index Md is 0.1, and the abnormal fluctuation index Yc is 5.0.

[0149] Set the threshold as follows:

[0150] Behavior standard threshold Q1 = 7.0;

[0151] Event safety threshold Q2 = 0.2;

[0152] Abnormal safety threshold Q3 = 4.0;

[0153] The comparison results are as follows:

[0154] The behavior index Xw < the behavior standard threshold Q1, indicating that the claimant's behavior is normal.

[0155] The event contradiction index Md < event safety threshold Q2, indicating that there is no obvious contradiction in the claim event.

[0156] The abnormal fluctuation index Yc < abnormal safety threshold Q3, indicating that the fluctuation of claim amount is normal.

[0157] Therefore, based on these comparison results, this claim case is considered normal and no additional processing is required.

[0158] Example 2:

[0159] F = 2 (the frequency value of the number of claims in the past year)

[0160] A=8000; Nt=3; E=1200; Cd=4; Td=30; At=400; R=60; C1=0.05

[0161] Substituting the above values ​​into the formula:

[0162] Behavior index Xw = (2 + 8000) / (3 + 1200) + 0.05 = 6.67;

[0163] Event contradiction index Md = 4 / 30 = 0.133;

[0164] Abnormal fluctuation index Yc = (400-60) / 60 = 5.67;

[0165] In this example, the behavior index Xw is 6.67, the event contradiction index Md is 0.133, and the abnormal fluctuation index Yc is 5.67.

[0166] According to the preset thresholds in Example 1, a comparison is made, the behavior index Xw=6.67, and the behavior standard threshold Q1=7.0. Result: the behavior index Xw<behavior standard threshold Q1, indicating that the claimant's behavior is normal.

[0167] Event contradiction index Md = 0.133, event safety threshold Q2 = 0.2 Result: Event contradiction index Md < event safety threshold Q2, indicating that there is no obvious contradiction in the claim event.

[0168] Abnormal fluctuation index Yc=5.67, abnormal safety threshold Q3=4.0 Result: Abnormal fluctuation index Yc>abnormal safety threshold Q3, indicating that the claim amount fluctuates abnormally.

[0169] According to these comparison results, the behavior of the claim case is normal, there are no obvious contradictions in the events, but the fluctuation of the claim amount is abnormal. Therefore, it is necessary to further review the claim case to investigate the abnormal fluctuation issue.

[0170] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A blockchain-based information processing method, characterized in that: The following steps are involved: S1. Establish an insurance claim blockchain network, set up a blockchain consensus mechanism, make several nodes in the blockchain have a copy of the complete blockchain, and establish claim rules, conditions and payment contracts; and use it to verify and analyze claim data; S2. Collecting historical claim record information submitted by the claimant and the claimant’s current claim information; S3. Feature extraction is performed on historical claim record information, and the claim frequency, historical claim amount, and event description are analyzed to determine whether there are any contradictions. Time series analysis methods are used to analyze the change in claim amount over time and whether there are any abnormal fluctuations. A claim model is established. Based on historical fraud cases and normal claim cases, the claimant's current claim information is analyzed and calculated to obtain a behavior index Xw, an event contradiction index Md, and an abnormal fluctuation index Yc. The behavior index Xw, event contradiction index Md, and abnormal fluctuation index Yc are obtained by the following formula: In the formula, F represents the frequency value of the number of claims made by the claimant in the past year, A represents the total amount of historical claims made by the claimant, Nt represents the number of different insurance types, E represents the expected value or average value of the insurance amount based on historical data; Cd represents the number of contradictory words or contradictory information in the claim document, Td represents the total number of words in the claim document; At represents the floating value of the claim amount changing with the event, R represents the expected normal fluctuation range calculated based on historical data; C1 represents the correction constant; S4, comparing the claimant's behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result; S5. Review the claimant’s current claim information based on the forecast assessment results.

2. The information processing method based on blockchain according to claim 1, characterized in that: Said S1 specifically includes: establishing a blockchain platform of Ethereum, Hyperledger, Fabric or Corda, setting a data structure and storage method of claim data, and configuring blockchain nodes, storing a copy of the complete blockchain in the blockchain nodes; And choose Proof of Work, Proof of Stake or Byzantine Fault Tolerant consensus mechanism to ensure the consistency of claim data between nodes; According to the needs of insurance claims, write smart contracts, perform functional tests, performance tests, and security tests on the smart contracts, and deploy the smart contracts to the blockchain nodes to ensure that the smart contracts are consistent with the claim rules and conditions.

3. The information processing method based on blockchain according to claim 1, characterized in that: The historical claim record information includes claimant personal information, historical claim type information, historical claim amount information, event date and event description information, claim processing result information and payment record information; The claimant's current claim information includes event description information, current claim amount information, claim event date and event additional information, current claim document information and claim status information. The claim document information includes medical certificates, accident reports and event photos.

4. The information processing method based on blockchain according to claim 1, characterized in that: Establish a claim model, extract and analyze features based on the claimant's historical claim record information, calculate the claim frequency F, the number of claims submitted for events in the past year, use the time window method to calculate, determine the starting point of the time window, slide the window according to the selected time window size, use counting operations to count the number of claims submitted by the claimant during the window period, and update it once a month; Calculate the average and standard deviation of the claimant's historical claim amounts to collect relevant indicators and obtain the distribution of the claimant's claim amounts; Use natural language processing (NLP) technology, including text mining and text analysis, to analyze whether there are any contradictions or anomalies in the description of historical claims, and make statistics; According to the time series analysis method, moving average or exponential smoothing method is used to analyze the trend of claim amount over time; Detect whether there is abnormal fluctuation and the value of abnormal fluctuation.

5. The information processing method based on blockchain according to claim 1, characterized in that: The S4 specifically includes: The behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant are respectively compared with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 to obtain the following prediction evaluation results; The predicted evaluation results include normal results, event-contradictory results, abnormal fluctuation results, and multiple problem results; The normal results include the following: If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal; If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event; If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal; When all indices are below the threshold, the claim is considered normal and no additional processing is required.

6. The information processing method based on blockchain according to claim 5 is characterized in that: The conflicting results of the incident include: If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal; If the event contradiction index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required; If the abnormal fluctuation index Yc < abnormal safety threshold Q3, it means that the fluctuation of claim amount is normal; Because the contradiction index Md exceeds the event safety threshold Q2, the claim requires additional review to resolve the contradiction.

7. The information processing method based on blockchain according to claim 5 is characterized in that: The abnormal fluctuation results include: If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal; If the event contradiction index Md < event safety threshold Q2, it means that there is no obvious contradiction in the claim event; If the abnormal fluctuation index Yc ≥ abnormal safety threshold Q3, it means that the claim amount fluctuates abnormally; Because the abnormal volatility index Yc exceeds the abnormal safety threshold Q3, the claim requires additional review to investigate abnormal volatility.

8. The information processing method based on blockchain according to claim 5 is characterized in that: The multiple question results include: If the behavior index Xw < the behavior standard threshold Q1, it means that the claimant’s behavior is normal; If the event contradiction index Md ≥ event safety threshold Q2, it means that there is a contradiction in the claim event and additional review is required; Abnormal fluctuation index Yc ≥ abnormal safety threshold Q3, indicating abnormal fluctuation of claim amount; Since both the contradiction index Md and the abnormal fluctuation index Yc exceed the threshold, the claim requires detailed review to resolve the contradiction and abnormal fluctuation issues.

9. The information processing method based on blockchain according to claim 1, characterized in that: According to the prediction evaluation results, the claimant's current claim information is considered normal, without abnormal behavior or contradictions, so the claim will continue to be processed and approved for payment; the claim information normal result indicates that the claimant's historical behavior is consistent with the current claim information, and no warning information needs to be triggered; If the predictive assessment results indicate potential abnormal behavior, inconsistencies, or risks, but it is not yet certain whether the claim is normal, the claim information will be marked as requiring additional review and marked as a fraud alert, triggering a warning message.

10. A blockchain-based information processing system, applied to a blockchain-based information processing method according to any one of claims 1 to 9, characterized in that: It includes blockchain network module, data collection module, feature extraction module, analysis module, behavior assessment module and review module; The blockchain network module is used to establish an insurance claim blockchain network, configure a blockchain consensus mechanism, ensure that nodes in the blockchain have a copy of the complete blockchain; and deploy smart contracts to manage the execution of claim data and rules; The data collection module is used to collect historical claim record information and current claim information submitted by the claimant, and adopts a data interface mode and a front-end interface interaction mode for the claimant to submit information; The feature extraction module is used to extract features from historical claim record information, including claim frequency, historical claim amount, and whether there is any contradiction in event description; The analysis module is used to analyze the change of claim amount over time using time series analysis method, detect abnormal fluctuations, and establish a claim model for fraud detection and analysis; The behavior evaluation module is used to calculate the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc of the claimant, and compare the calculated behavior index Xw, event contradiction index Md and abnormal fluctuation index Yc with the behavior standard threshold Q1, event safety threshold Q2 and abnormal safety threshold Q3 respectively to obtain a prediction evaluation result; The review module is used to review the claimant's current claim information based on the prediction and evaluation results, which include normal results, event-contradictory results, abnormal fluctuation results, and multiple-problem results.