Insurance case shunting method, system, device and medium
By using an insurance case triage rule engine and a neural network to identify compensation ratio parameters, the problem of rigid rules in the traditional insurance claims model has been solved, enabling accurate calculation of complex cases and efficient resource allocation, thereby improving the efficiency and accuracy of insurance claims processing.
Patent Information
- Application Number
- CN202511859392.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-10
- Publication Date
- 2026-02-27
AI Technical Summary
Traditional insurance claims processing models are ill-suited to the needs of industry development. Rule engines suffer from rigid rules, making it difficult to accurately identify payout ratios for complex cases, leading to uneven resource allocation and claims disputes.
An insurance case triage method is adopted, which performs preliminary classification through an insurance case triage rule engine, identifies complex cases, builds a recent historical case database, uses a compensation ratio parameter identification neural network to generate accurate compensation ratio parameters, and generates a claims settlement report in combination with applicable insurance rules.
It enables accurate claims settlement for complex cases, improves the efficiency and accuracy of claims processing, avoids prediction errors caused by outdated rules, ensures the compliance and accuracy of the processing results, and improves resource utilization efficiency and customer satisfaction.
Smart Images
Figure CN121582012A_ABST
Abstract
Description
Technical Field
[0001] This application pertains to the field of insurance triage, and in particular relates to a method, system, equipment, and medium for triaging insurance cases. Background Technology
[0002] The efficiency and accuracy of insurance claims processing often determine an insurance company's service capabilities and market competitiveness. With the rapid expansion of the insurance industry, the number of insurance cases is growing exponentially. At the same time, the emergence of new risk scenarios has made the structure of insurance cases increasingly complex. Traditional claims processing models are no longer able to meet the needs of industry development, driving the continuous iteration and evolution of insurance case diversion technology.
[0003] In the early days, insurance case handling relied heavily on manual processes. From entering case information and verifying documents to determining the type of claim and reviewing claims, each step required manual handling by claims personnel. There was a lack of uniform standards for classifying case types and grasping compensation standards, which not only led to an imbalance in the allocation of resources between ordinary and complex cases, but also easily caused claims disputes due to judgment errors.
[0004] To overcome the bottlenecks of manual processing, rule engines have been gradually applied to the insurance claims field. Insurance companies transform insurance contract terms, regulatory rules, and claims standards into structured rules to determine case types based on case information. However, with the further development of the insurance business, the limitations of rule engines have become increasingly apparent. Rule engines often suffer from rule rigidity, leading to insufficient adaptability to new types of cases. Especially when facing scenarios involving special drug use and complex medical procedures, pre-set rules are prone to coverage blind spots, failing to accurately identify payout ratios and hindering accurate calculation of complex cases. Summary of the Invention
[0005] Therefore, it is necessary to provide an insurance case triage method, system, device, and medium that can avoid the rigidity of rule engines and achieve accurate claims settlement for complex cases, in order to address the above-mentioned technical problems.
[0006] Firstly, this application provides a method for diverting insurance cases, including:
[0007] The system obtains insurance case information and inputs it into the insurance case triage rule engine to obtain case type information and applicable insurance rules. Case type information includes information on ordinary case types and information on case types with payout ratio restrictions.
[0008] Insurance cases whose case type information is a payout ratio limit case type information are identified as complex insurance cases, and the payout ratio limit rules and payout ratio parameters of the applicable insurance rules in the complex insurance cases are obtained;
[0009] If the rule engine update interval of the insurance case triage rule engine exceeds the preset rule engine degradation cycle threshold, recent historical cases that are within the rule engine update cycle and real-time case analysis cycle are obtained from the historical case library, and a recent historical case library is constructed based on the recent historical cases.
[0010] Based on the rules on the payout ratio limit for complex insurance cases, similar cases of complex insurance cases are identified from the recent historical case database;
[0011] The insurance case information of complex insurance cases, the insurance case information of similar cases, and the actual payout ratio parameters of similar cases are input into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of complex insurance cases corresponding to the payout ratio parameters.
[0012] Based on recent historical payout ratios and applicable insurance rules, a case triage and settlement report is generated for complex insurance cases.
[0013] Secondly, this application also provides an insurance case triage system, including:
[0014] The initial insurance case triage module is used to obtain insurance case information and input it into the insurance case triage rule engine to obtain the case type information and applicable insurance rules. The case type information includes ordinary case type information and compensation ratio restriction case type information.
[0015] The complex insurance case identification module is used to identify insurance cases whose case type information is a payout ratio restriction case type information as complex insurance cases, and to obtain the payout ratio restriction rule and the payout ratio parameter of the payout ratio restriction rule in the applicable insurance rules for complex insurance cases;
[0016] The recent historical case identification module is used to obtain recent historical cases that are within the rule engine update cycle and real-time case analysis cycle from the historical case library if the rule engine update interval of the insurance case diversion rule engine exceeds the preset rule engine deterioration cycle threshold, and to build a recent historical case library based on the recent historical cases.
[0017] The similar insurance case identification module is used to identify similar cases of complex insurance cases from the recent historical case database based on the compensation ratio restriction rules for complex insurance cases.
[0018] The payout ratio parameter calculation module is used to input the insurance case information of complex insurance cases, the insurance case information of similar cases, and the actual payout ratio parameters of similar cases into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of complex insurance cases corresponding to the payout ratio parameters.
[0019] The claims settlement report generation module is used to generate claims settlement reports for complex insurance cases based on recent historical similar claims ratio parameters and applicable insurance rules.
[0020] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method as described in any of the first aspects of this application.
[0021] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in any of the first aspects of this application.
[0022] The aforementioned insurance case triage methods, systems, equipment, and media, through an insurance case triage rule engine, can automatically identify and classify case types, dynamically match case type information corresponding to insurance cases, improve the accuracy of case classification, and enhance the efficiency of insurance case processing. By specializing in the processing of complex cases, more computing resources and stricter compliance verification processes are allocated to them, ensuring that processing results meet business rule requirements and avoiding discrepancies in payout ratios due to rule lag. This not only significantly improves resource utilization efficiency but also guarantees the compliance and accuracy of handling high-risk cases. By constructing a dynamically updated recent historical case database, data support is provided for decision-making in complex cases, compensating for data gaps and rule rigidity caused by rule engine update lags, avoiding prediction biases due to outdated rules, and maintaining stable decision-making accuracy. Through a payout ratio parameter identification neural network, recent historical payout ratio parameters conforming to historical data distribution are generated, and combined with applicable insurance rules to generate claims reports. This significantly improves the accuracy and stability of payout ratio predictions, reduces the error rate of payout ratio predictions, and thus provides accurate payout ratio parameters for claims decisions in complex cases. Attached Figure Description
[0023] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0024] Figure 1 A flowchart illustrating an insurance case triage method provided in one embodiment of this application. Figure 1 ;
[0025] Figure 2A flowchart illustrating an insurance case triage method provided in one embodiment of this application. Figure 2 ;
[0026] Figure 3 This is a schematic diagram of the structure of an insurance case triage system provided in one embodiment of this application. Detailed Implementation
[0027] 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.
[0028] In one exemplary embodiment of this application, such as Figure 1 As shown, an insurance case triage method is provided. This embodiment illustrates the application of this method to an insurance case triage terminal. It is understood that this method can also be applied to an insurance case triage server, and further to an insurance case triage system including both an insurance case triage terminal and an insurance case triage server, and is implemented through the interaction between the two. In this embodiment, the method may include the following steps:
[0029] Step S101: Obtain insurance case information and input the insurance case information into the insurance case diversion rule engine to obtain the case type information and applicable insurance rules of the insurance case.
[0030] Specifically, the insurance case triage terminal can obtain insurance case information and input the obtained insurance case information into the insurance case triage rule engine installed on the terminal to perform insurance case triage rule matching, thereby obtaining the case type information and applicable insurance rules of the insurance case.
[0031] Optionally, case type information may include, but is not limited to, information on ordinary case types and information on case types with payout ratio restrictions.
[0032] As an illustration, the insurance case triage rule engine can be installed on the insurance case triage terminal, the insurance case triage server, or the insurance case triage system; no specific limitation is made here.
[0033] For example, the information on case types with reimbursement ratio restrictions may include, but is not limited to, one or more of the following: information on reimbursement ratio restrictions for consistency of pathological medication, information on reimbursement ratio restrictions for special drugs, information on reimbursement ratio restrictions for treatment items, and information on reimbursement ratio restrictions for medical institution levels. Applicable insurance rules may include, but are not limited to, general insurance rules and reimbursement ratio restriction rules. Specifically, the reimbursement ratio restriction rules may include, but are not limited to, one or more of the following: reimbursement ratio restriction rules for consistency of pathological medication; reimbursement ratio restriction rules for special drugs; reimbursement ratio restriction rules for treatment items; and reimbursement ratio restriction rules for medical institution levels.
[0034] Optionally, insurance case information may include, but is not limited to, disease diagnosis information, treatment plan information, drug list information, medical institution information, and medical expense information.
[0035] Optionally, the insurance case triage terminal can, based on the insurance case triage rule engine, set the case type information of insurance cases with applicable insurance rules that include a payout ratio restriction as payout ratio restriction case type information.
[0036] Furthermore, the insurance case triage terminal can, based on the insurance case triage rule engine, set the case type information of insurance cases for which there are no payout ratio restrictions in the applicable insurance rules as ordinary case type information.
[0037] Step S102: Identify insurance cases whose case type information is a payout ratio restriction case type information as complex insurance cases, and obtain the payout ratio restriction rule and the payout ratio parameter of the payout ratio restriction rule in the applicable insurance rules for complex insurance cases.
[0038] Specifically, the insurance case triage terminal can identify insurance cases with payout ratio restriction information as complex insurance cases. It can also identify insurance cases with ordinary case information as ordinary insurance cases. Furthermore, the terminal can obtain the payout ratio restriction rules from the applicable insurance rules for complex insurance cases and retrieve the payout ratio parameters corresponding to the pre-set payout ratio restriction rules based on the insurance case triage rule engine.
[0039] Optionally, if the rule engine update interval of the insurance case triage rule engine is less than the rule engine degradation cycle threshold, an insurance case triage settlement report for complex insurance cases will be generated based on the payout ratio parameter and applicable insurance rules, combined with insurance case information.
[0040] Furthermore, the insurance case triage terminal can generate an insurance case triage and settlement report for ordinary insurance cases based on the payout ratio parameters corresponding to the applicable insurance rules for ordinary insurance cases, combined with the applicable insurance rules and insurance case information.
[0041] Step S103: If the rule engine update interval of the insurance case diversion rule engine exceeds the preset rule engine degradation cycle threshold, recent historical cases within the rule engine update cycle and real-time case analysis cycle are obtained from the historical case library, and a recent historical case library is constructed based on the recent historical cases.
[0042] Specifically, the insurance case triage terminal can obtain the rule engine update interval of the insurance case triage rule engine when insurance case information is input into the insurance case triage rule engine equipped on the insurance case triage terminal for insurance case triage rule matching. If the rule engine update interval of the insurance case triage rule engine exceeds the preset rule engine degradation cycle threshold, the insurance case triage terminal can obtain recent historical cases from the historical case database that are within the current rule engine update cycle and real-time case analysis cycle of the insurance case triage rule engine, and build a recent historical case database based on the recent historical cases.
[0043] Optionally, the current rule engine update cycle of the insurance case triage rule engine can start from the last insurance case triage rule engine update and end when the insurance case information is input into the insurance case triage rule engine equipped on the insurance case triage terminal for insurance case triage rule matching.
[0044] Optionally, the real-time case analysis period can be used to characterize the time boundary corresponding to recent historical cases that are of reference value.
[0045] Step S104: Based on the rules for limiting the payout ratio of complex insurance cases, identify similar cases of complex insurance cases from the recent historical case database.
[0046] Specifically, the insurance case triage terminal can identify recent historical cases in the recent historical case database that have the same payout ratio limit rules as complex insurance cases as similar cases to complex insurance cases, based on the payout ratio limit rules for complex insurance cases.
[0047] Step S105: Input the insurance case information of complex insurance cases, the insurance case information of each similar case, and the actual payout ratio parameters of each similar case into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of complex insurance cases corresponding to the payout ratio parameters.
[0048] Specifically, the insurance case triage terminal can input the insurance case information of complex insurance cases, the insurance case information of each similar case of complex insurance cases, and the actual payout ratio parameters of each similar case of complex insurance cases into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of complex insurance cases corresponding to the payout ratio parameters.
[0049] As an illustration, the neural network for recognizing the compensation ratio parameter can be mounted on an insurance case triage terminal, an insurance case triage server, or an insurance case triage system; no specific limitations are made here.
[0050] As an illustration, the neural network for identifying the payout ratio parameter can be constructed based on, but is not limited to, one or more neural networks such as convolutional neural networks, recurrent neural networks, multilayer perceptrons, Transformer networks, and attention mechanism networks.
[0051] Step S106: Based on recent historical similar payout ratio parameters and applicable insurance rules, generate an insurance case triage and settlement report for complex insurance cases.
[0052] In the aforementioned insurance case triage method, the insurance case triage rule engine can accurately classify insurance case types, thereby identifying complex insurance cases with payout ratio restrictions, improving the efficiency of insurance case settlement, and reducing the deviation of insurance case settlement results. By dynamically monitoring the update status of the rule engine and building a recent historical case database, it can dynamically adapt to changes and developments in insurance business scenarios, effectively avoiding triage errors caused by the rigidity of the rule engine, and improving the adaptability of insurance case triage. By intelligently calculating payout ratio parameters, it can leverage the feature extraction and data processing capabilities of neural networks to deeply explore the potential correlations between complex cases and similar cases, thereby improving the scientific nature and adaptability of payout ratio parameters for complex cases, and enhancing the accuracy of claims processing and customer satisfaction.
[0053] In an optional embodiment of this application, please refer to Figure 2 Other methods for diverting insurance cases may include:
[0054] Step S211: If the rule engine update interval of the insurance case diversion rule engine is less than the rule engine deterioration cycle threshold, an insurance case diversion settlement report for complex insurance cases is generated based on the payout ratio parameter and the applicable insurance rules.
[0055] In an optional embodiment of this application, the insurance case information may include disease diagnosis information, treatment plan information, drug list information, medical institution information, and medical expense information, and the insurance case triage rule engine may include a disease association rule engine, an expense association rule engine, and an institution level association rule engine.
[0056] Please refer to Figure 1 and Figure 2 Step S101 involves obtaining insurance case information and inputting it into the insurance case triage rule engine to identify the case type and applicable insurance rules, which may include:
[0057] Step S201: Obtain insurance case information and perform structured processing on the insurance case information to obtain standard fields for insurance case information.
[0058] Step S202: Match the standard fields of insurance case information with the disease association rule engine, the cost association rule engine, and the institution level association rule engine to obtain the applicable insurance rules for the insurance case.
[0059] Step S203: If the applicable insurance rules for the insurance case contain a payout ratio restriction rule, set the case type information of the insurance case to payout ratio restriction case type information; otherwise, set the case type information of the insurance case to ordinary case type information.
[0060] In an optional embodiment of this application, the standard fields of insurance case information are matched with the disease association rule engine, the cost association rule engine, and the institution level association rule engine to obtain the applicable insurance rules for the insurance case, which may include:
[0061] Specifically, the insurance case triage terminal can calculate the rule matching degree between the standard fields of insurance case information and each disease association rule in the disease association rule engine, and construct the applicable insurance rules for the insurance case based on the disease association rules whose rule matching degree is greater than the preset rule matching threshold of each disease association rule.
[0062] Specifically, the insurance case triage terminal can calculate the rule matching degree between the standard fields of insurance case information and each cost association rule in the cost association rule engine, and construct the applicable insurance rules for the insurance case based on the cost association rules whose rule matching degree is greater than the preset rule matching threshold of each cost association rule.
[0063] Specifically, the insurance case triage terminal can calculate the rule matching degree between the standard fields of insurance case information and the various institutional level association rules in the institutional level association rule engine, and construct the applicable insurance rules for the insurance case based on the institutional level association rules whose rule matching degree is greater than the preset rule matching threshold corresponding to each institutional level association rule.
[0064] Optionally, the expression for rule matching degree can be:
[0065]
[0066] In the formula, For insurance cases Same as the first Insurance Case Triage Rules Rule matching degree, For the first Insurance Case Triage Rules The number of feature dimensions, For the first Insurance Case Triage Rules The Feature dimension weight coefficients of each feature dimension For the first Insurance Case Triage Rules The Feature dimensions are the same as those of insurance cases The single-dimensional matching degree.
[0067] In an optional embodiment of this application, the information on the type of case for which the reimbursement ratio is limited may include information on the type of reimbursement ratio for consistency of pathological medication, information on the type of reimbursement ratio for special drugs, information on the type of reimbursement ratio for treatment items, and information on the type of reimbursement ratio for medical institution level. The reimbursement ratio limitation rules may include the reimbursement ratio limitation rules for consistency of pathological medication corresponding to the disease association rule engine, the reimbursement ratio limitation rules for special drugs corresponding to the cost association rule engine, the reimbursement ratio limitation rules for treatment items corresponding to the cost association rule engine, and the reimbursement ratio limitation rules for medical institution level corresponding to the institution level association rule engine.
[0068] According to the rules on payout ratio limits for complex insurance cases, similar cases of complex insurance cases identified from the recent historical case database can include:
[0069] Specifically, if the payout ratio restriction rules for complex insurance cases include the payout ratio restriction rules for consistency of pathological medication, the insurance case diversion terminal can identify recent historical cases in the recent historical case database that apply insurance rules including the payout ratio restriction rules for consistency of pathological medication as similar cases of payout ratio restriction for consistency of pathological medication, and construct similar cases based on similar cases of payout ratio restriction for consistency of pathological medication.
[0070] Specifically, if the reimbursement ratio restriction rules for complex insurance cases include the reimbursement ratio restriction rules for special drugs, the insurance case diversion terminal can identify recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio restriction rules for special drugs as similar cases of reimbursement ratio restriction for special drugs, and construct similar cases based on similar cases of reimbursement ratio restriction for special drugs.
[0071] Specifically, if the reimbursement ratio limit rules for complex insurance cases include the reimbursement ratio limit rules for medical treatment items, the insurance case diversion terminal can identify recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio limit rules for medical treatment items as similar cases with reimbursement ratio limits for medical treatment items, and construct similar cases based on similar cases with reimbursement ratio limits for medical treatment items.
[0072] Specifically, if the reimbursement ratio limit rules for complex insurance cases include the reimbursement ratio limit rules for medical institution levels, the insurance case diversion terminal can identify recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio limit rules for medical institution levels as similar cases of reimbursement ratio limit for medical institution levels, and construct similar cases based on similar cases of reimbursement ratio limit for medical institution levels.
[0073] In an optional embodiment of this application, please refer to Figure 1 and Figure 2 Step S105 involves inputting the insurance case information of complex insurance cases, the insurance case information of similar cases, and the actual payout ratio parameters of each similar case into the payout ratio parameter recognition neural network to generate recent historical similar payout ratio parameters for complex insurance cases corresponding to the payout ratio parameters. These parameters may include:
[0074] Step S207: Input the insurance case information of complex insurance cases and the insurance case information of similar cases into the insurance case information feature extraction component in the compensation ratio parameter recognition neural network to obtain the insurance case features of complex insurance cases and the insurance case features of similar cases.
[0075] Step S208: Input the insurance case characteristics of complex insurance cases and the insurance case characteristics of each similar case into the insurance association weight extraction component in the compensation ratio parameter recognition neural network, and calculate the insurance association weight of each similar case corresponding to the complex insurance case.
[0076] Step S209: Input the insurance association weights of each similar case and the actual payout ratio parameters of each similar case into the global feature analysis component in the payout ratio parameter recognition neural network to generate recent historical similar payout ratio parameters for complex insurance cases.
[0077] In an optional embodiment of this application, the expression for the loss function of the compensation ratio parameter identification neural network can be:
[0078]
[0079]
[0080]
[0081]
[0082]
[0083] In the formula, The loss function of the neural network is used to identify the payout ratio parameter. , , and These represent the network output predicted loss coefficient, the network output predicted loss term, the weighted predicted loss coefficient, and the weighted predicted loss term, respectively, while represents the total number of training samples for complex insurance cases. For the first The payout ratio parameter identification neural network outputs recent historical payout ratio parameters of similar cases from training samples of complex insurance cases. For the first Training labels for the actual payout ratio of a training sample of complex insurance cases. For the first The KL divergence between the insurance association weight distribution and the true weight training label distribution of a complex insurance case training sample. and The first The insurance association weight distribution and the true weight training label distribution of a complex insurance case training sample. For the first The total number of similar cases corresponding to the training sample of complex insurance cases. and The first The training sample for complex insurance cases is the same as the first one. The training sample corresponding to the complex insurance case The insurance association weights and true weights among similar cases were used to train the labels. The parameter is the minimum non-zero parameter in the denominator.
[0084] In one exemplary embodiment of this application, such as Figure 2 As shown, an insurance case triage method is provided, including:
[0085] Step S201: Obtain insurance case information and perform structured processing on the insurance case information to obtain standard fields for insurance case information.
[0086] Step S202: Match the standard fields of insurance case information with the disease association rule engine, the cost association rule engine, and the institution level association rule engine to obtain the applicable insurance rules for the insurance case.
[0087] Step S203: If the applicable insurance rules for the insurance case contain a payout ratio restriction rule, set the case type information of the insurance case to payout ratio restriction case type information; otherwise, set the case type information of the insurance case to ordinary case type information.
[0088] Step S204: Identify insurance cases whose case type information is a payout ratio restriction case type information as complex insurance cases, and obtain the payout ratio restriction rule and the payout ratio parameter of the payout ratio restriction rule in the applicable insurance rules for complex insurance cases.
[0089] Step S205: If the rule engine update interval of the insurance case diversion rule engine exceeds the preset rule engine degradation cycle threshold, recent historical cases within the rule engine update cycle and real-time case analysis cycle are obtained from the historical case library, and a recent historical case library is constructed based on the recent historical cases.
[0090] Step S206: Based on the rules for limiting the payout ratio of complex insurance cases, identify similar cases of complex insurance cases from the recent historical case database.
[0091] Step S207: Input the insurance case information of complex insurance cases and the insurance case information of similar cases into the insurance case information feature extraction component in the compensation ratio parameter recognition neural network to obtain the insurance case features of complex insurance cases and the insurance case features of similar cases.
[0092] Step S208: Input the insurance case characteristics of complex insurance cases and the insurance case characteristics of each similar case into the insurance association weight extraction component in the compensation ratio parameter recognition neural network, and calculate the insurance association weight of each similar case corresponding to the complex insurance case.
[0093] Step S209: Input the insurance association weights of each similar case and the actual payout ratio parameters of each similar case into the global feature analysis component in the payout ratio parameter recognition neural network to generate recent historical similar payout ratio parameters for complex insurance cases.
[0094] Step S210: Based on recent historical similar payout ratio parameters and applicable insurance rules, generate an insurance case triage and settlement report for complex insurance cases.
[0095] Step S211: If the rule engine update interval of the insurance case diversion rule engine is less than the rule engine deterioration cycle threshold, an insurance case diversion settlement report for complex insurance cases is generated based on the payout ratio parameter and the applicable insurance rules.
[0096] In the aforementioned insurance case triage method, a multi-dimensional rule engine is used to initially classify insurance cases, accurately identifying "cases with payout ratio restrictions" that require complex judgment. When the efficiency of the rule engine deteriorates over time, a dynamic adaptive mechanism is activated to build a dynamic knowledge base from recent historical cases. Combined with a neural network that identifies payout ratio parameters, the system analyzes the correlation between current complex cases and similar historical cases to generate accurate and timely suggested payout ratios. This significantly improves the intelligence and efficiency of the insurance claims process, effectively enhances the accuracy and consistency of claims calculations for complex cases, and ultimately achieves optimized allocation of claims resources and efficient management of claims risks.
[0097] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.
[0098] Based on the same inventive concept, this application also provides an insurance case triage system for implementing the insurance case triage method described above. The solution provided by this system is similar to the solution described in the above method; therefore, the specific limitations of one or more insurance case triage system embodiments provided below can be found in the limitations of the insurance case triage method described above, and will not be repeated here.
[0099] In one exemplary embodiment, such as Figure 3 As shown, an insurance case triage system 300 is provided, including:
[0100] The initial insurance case triage module 301 can be used to obtain insurance case information and input the insurance case triage rule engine to obtain the case type information and applicable insurance rules of the insurance case. The case type information can include ordinary case type information and compensation ratio restriction case type information.
[0101] The complex insurance case identification module 302 can be used to identify insurance cases whose case type information is a payout ratio restriction case type information as complex insurance cases, and obtain the payout ratio restriction rule and the payout ratio parameter of the payout ratio restriction rule in the applicable insurance rules of complex insurance cases.
[0102] The recent historical case identification module 303 can be used to obtain recent historical cases that are within the rule engine update cycle and real-time case analysis cycle from the historical case library if the rule engine update interval of the insurance case diversion rule engine exceeds the preset rule engine degradation cycle threshold, and construct a recent historical case library based on the recent historical cases.
[0103] The similar insurance case identification module 304 can be used to identify similar cases of complex insurance cases from the recent historical case database based on the compensation ratio restriction rules for complex insurance cases.
[0104] The payout ratio parameter calculation module 305 can be used to input the insurance case information of complex insurance cases, the insurance case information of similar cases, and the actual payout ratio parameters of similar cases into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of complex insurance cases corresponding to the payout ratio parameters.
[0105] The claims settlement report generation module 306 can be used to generate claims settlement reports for complex insurance cases based on recent historical similar claims ratio parameters and applicable insurance rules.
[0106] In an optional embodiment of this application, the insurance case triage system 300 can also be used for:
[0107] If the rule engine update interval of the insurance case triage rule engine is less than the rule engine degradation cycle threshold, an insurance case triage and settlement report for complex insurance cases will be generated based on the payout ratio parameter and the applicable insurance rules.
[0108] In an optional embodiment of this application, the insurance case preliminary triage module 301 can also be used for:
[0109] Obtain insurance case information and perform structured processing on the insurance case information to obtain standard fields for insurance case information.
[0110] The standard fields of insurance case information are matched with the disease association rule engine, the cost association rule engine, and the institution level association rule engine to obtain the applicable insurance rules for the insurance case.
[0111] If the applicable insurance rules for an insurance case contain a payout ratio restriction rule, the case type information of the insurance case will be set to the payout ratio restriction case type information; otherwise, the case type information of the insurance case will be set to the ordinary case type information.
[0112] In an optional embodiment of this application, the insurance case preliminary triage module 301 can also be used for:
[0113] The system calculates the rule matching degree between the standard fields of insurance case information and each disease association rule in the disease association rule engine, and constructs applicable insurance rules for insurance cases based on disease association rules whose rule matching degree is greater than the preset rule matching threshold for each disease association rule.
[0114] Calculate the rule matching degree between the standard fields of insurance case information and each cost association rule in the cost association rule engine, and construct the applicable insurance rules for the insurance case based on the cost association rules whose rule matching degree is greater than the preset rule matching threshold of each cost association rule.
[0115] The system calculates the rule matching degree between the standard fields of insurance case information and the rule matching rules of each institution level in the institution level association rule engine, and constructs the applicable insurance rules for the insurance case based on the institution level association rules whose rule matching degree is greater than the preset rule matching threshold of each institution level association rule.
[0116] In an optional embodiment of this application, the similar insurance case identification module 304 can also be used for:
[0117] If the reimbursement ratio restriction rules for complex insurance cases include the reimbursement ratio restriction rules for consistency of pathological medication, then recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio restriction rules for consistency of pathological medication will be identified as similar cases of reimbursement ratio restriction for consistency of pathological medication, and similar cases will be constructed based on similar cases of reimbursement ratio restriction for consistency of pathological medication.
[0118] If the reimbursement ratio limit rules for complex insurance cases include special drug reimbursement ratio limit rules, then recent historical cases in the recent historical case database that apply insurance rules including special drug reimbursement ratio limit rules will be identified as similar cases of special drug reimbursement ratio limit, and similar cases will be constructed based on similar cases of special drug reimbursement ratio limit.
[0119] If the reimbursement ratio limit rules for complex insurance cases include the reimbursement ratio limit rules for medical treatment items, then recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio limit rules for medical treatment items will be identified as similar cases of reimbursement ratio limit for medical treatment items, and similar cases will be constructed based on similar cases of reimbursement ratio limit for medical treatment items.
[0120] If the reimbursement ratio limit rules for complex insurance cases include the reimbursement ratio limit rules for medical institution levels, then recent historical cases in the recent historical case database that apply insurance rules including the reimbursement ratio limit rules for medical institution levels will be identified as similar cases of reimbursement ratio limit for medical institution levels, and similar cases will be constructed based on similar cases of reimbursement ratio limit for medical institution levels.
[0121] In an optional embodiment of this application, the compensation ratio parameter calculation module 305 can also be used for:
[0122] The insurance case information of complex insurance cases and insurance case information of similar cases are input into the insurance case information feature extraction component in the compensation ratio parameter recognition neural network to obtain the insurance case features of complex insurance cases and insurance case features of similar cases.
[0123] The insurance case characteristics of complex insurance cases and the insurance case characteristics of similar cases are input into the insurance association weight extraction component in the compensation ratio parameter recognition neural network to calculate the insurance association weight of each similar case corresponding to the complex insurance case.
[0124] The insurance association weights of each similar case and the actual payout ratio parameters of each similar case are input into the global feature analysis component in the payout ratio parameter recognition neural network to generate recent historical similar payout ratio parameters for complex insurance cases.
[0125] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the insurance case triage method as described above.
[0126] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.
[0127] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and 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 modules can be selected to achieve the purpose of this disclosure according to actual needs. Those skilled in the art can understand and implement this without creative effort.
[0128] The above-described embodiments are merely illustrative of several implementation methods of the embodiments of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the embodiments of this application, and these modifications and improvements all fall within the protection scope of the embodiments of this application.
Claims
1. A method for diverting insurance cases, characterized in that, The method includes: The insurance case information is obtained and input into the insurance case diversion rule engine to obtain the case type information and applicable insurance rules of the insurance case. The case type information includes ordinary case type information and compensation ratio restriction case type information. The insurance case whose case type information is the compensation ratio limit case type information is identified as a complex insurance case, and the compensation ratio limit rule and the compensation ratio parameter of the compensation ratio limit rule in the applicable insurance rules of the complex insurance case are obtained; If the rule engine update interval of the insurance case triage rule engine exceeds the preset rule engine degradation cycle threshold, recent historical cases within the rule engine update cycle and real-time case analysis cycle are obtained from the historical case library, and a recent historical case library is constructed based on the recent historical cases. Based on the payout ratio restriction rules for the complex insurance case, similar cases to the complex insurance case are identified from the recent historical case database; The insurance case information of the complex insurance case, the insurance case information of each of the similar cases, and the actual payout ratio parameters of each of the similar cases are input into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of the complex insurance case corresponding to the payout ratio parameters; Based on the recent historical similar payout ratio parameters and the applicable insurance rules, an insurance case triage and settlement report for the complex insurance case is generated.
2. The method according to claim 1, characterized in that, The method further includes: If the rule engine update interval of the insurance case triage rule engine is less than the rule engine degradation cycle threshold, an insurance case triage settlement report for the complex insurance case is generated based on the payout ratio parameter and the applicable insurance rules.
3. The method according to claim 1, characterized in that, The insurance case information includes disease diagnosis information, treatment plan information, drug list information, medical institution information, and medical expense information. The insurance case triage rule engine includes a disease association rule engine, an expense association rule engine, and an institution level association rule engine. The process of obtaining insurance case information and inputting it into the insurance case triage rule engine to identify the case type information and applicable insurance rules includes: Obtain the insurance case information of the insurance case, and perform structured processing on the insurance case information to obtain standard fields for insurance case information; The standard fields of the insurance case information are matched with the disease association rule engine, the cost association rule engine and the institution level association rule engine to obtain the applicable insurance rules for the insurance case. If the applicable insurance rules for the insurance case contain the compensation ratio restriction rule, the case type information of the insurance case is set to the compensation ratio restriction case type information; otherwise, the case type information of the insurance case is set to the ordinary case type information.
4. The method according to claim 3, characterized in that, The step of matching the standard fields of the insurance case information with the disease association rule engine, the cost association rule engine, and the institution level association rule engine to obtain the applicable insurance rules for the insurance case includes: Calculate the rule matching degree between the standard field of the insurance case information and each disease association rule in the disease association rule engine, and construct the applicable insurance rule for the insurance case based on the disease association rule whose rule matching degree is greater than the preset rule matching threshold corresponding to each disease association rule; Calculate the rule matching degree between the standard field of the insurance case information and each cost association rule in the cost association rule engine, and construct the applicable insurance rule for the insurance case based on the cost association rule whose rule matching degree is greater than the preset rule matching threshold corresponding to each cost association rule; Calculate the rule matching degree between the standard field of the insurance case information and each institution level association rule in the institution level association rule engine, and construct the applicable insurance rule for the insurance case based on the institution level association rule whose rule matching degree is greater than the preset rule matching threshold corresponding to each institution level association rule; The expression for the rule matching degree is: In the formula, For the aforementioned insurance case Same as the first Insurance Case Triage Rules The matching degree of the aforementioned rules, For the first Insurance Case Triage Rules The number of feature dimensions, For the first Insurance Case Triage Rules The Feature dimension weight coefficients of each feature dimension For the first Insurance Case Triage Rules The The feature dimensions are the same as those of the insurance cases described. The single-dimensional matching degree.
5. The method according to claim 3, characterized in that, The information on the types of cases with limited reimbursement ratios includes information on limited reimbursement ratios for consistency of pathological medication, limited reimbursement ratios for special drugs, limited reimbursement ratios for treatment items, and limited reimbursement ratios for medical institution levels. The rules for limiting reimbursement ratios include rules for limited reimbursement ratios for consistency of pathological medications corresponding to the disease association rule engine, rules for limited reimbursement ratios for special drugs corresponding to the cost association rule engine, rules for limited reimbursement ratios for treatment items corresponding to the cost association rule engine, and rules for limited reimbursement ratios for medical institution levels corresponding to the institution level association rule engine. The step of identifying similar cases to the complex insurance case from the recent historical case database based on the payout ratio limitation rules for the complex insurance case includes: If the reimbursement ratio restriction rule for the complex insurance case includes the reimbursement ratio restriction rule for consistency of pathological medication, the recent historical cases in the recent historical case database whose applicable insurance rules include the reimbursement ratio restriction rule for consistency of pathological medication are identified as similar cases of reimbursement ratio restriction for consistency of pathological medication, and similar cases are constructed based on the similar cases of reimbursement ratio restriction for consistency of pathological medication. If the reimbursement ratio restriction rule for the complex insurance case includes the reimbursement ratio restriction rule for special drugs, the recent historical cases in the recent historical case database whose applicable insurance rules include the reimbursement ratio restriction rule for special drugs are identified as similar cases of reimbursement ratio restriction for special drugs, and similar cases are constructed based on the similar cases of reimbursement ratio restriction for special drugs. If the reimbursement ratio restriction rule for the complex insurance case includes the reimbursement ratio restriction rule for the medical treatment item, the recent historical cases in the recent historical case database whose applicable insurance rules include the reimbursement ratio restriction rule for the medical treatment item are identified as similar cases of reimbursement ratio restriction for the medical treatment item, and similar cases are constructed based on the similar cases of reimbursement ratio restriction for the medical treatment item. If the reimbursement ratio restriction rule for the complex insurance case includes the reimbursement ratio restriction rule for the medical institution level, the recent historical cases in the recent historical case database whose applicable insurance rules include the reimbursement ratio restriction rule for the medical institution level are identified as similar cases of reimbursement ratio restriction for the medical institution level, and similar cases are constructed based on the similar cases of reimbursement ratio restriction for the medical institution level.
6. The method according to any one of claims 1 to 5, characterized in that, The step of inputting the insurance case information of the complex insurance case, the insurance case information of each of the similar cases, and the actual payout ratio parameters of each of the similar cases into a payout ratio parameter recognition neural network to generate recent historical similar payout ratio parameters of the complex insurance case corresponding to the payout ratio parameters includes: The insurance case information of the complex insurance case and the insurance case information of each of the similar cases are input into the insurance case information feature extraction component in the compensation ratio parameter recognition neural network to obtain the insurance case features of the complex insurance case and the insurance case features of each of the similar cases. The insurance case features of the complex insurance case and the insurance case features of each of the similar cases are input into the insurance association weight extraction component in the compensation ratio parameter recognition neural network to calculate the insurance association weight of each of the similar cases corresponding to the complex insurance case. The insurance association weights of each of the aforementioned similar cases and the actual payout ratio parameters of each of the aforementioned similar cases are input into the global feature analysis component in the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of the complex insurance case.
7. The method according to claim 6, characterized in that, The expression for the loss function of the neural network for identifying the compensation ratio parameter is: In the formula, The loss function of the neural network is used to identify the compensation ratio parameter. , , and These represent the network output predicted loss coefficient, the network output predicted loss term, the weighted predicted loss coefficient, and the weighted predicted loss term, respectively, while represents the total number of training samples for complex insurance cases. For the first The payout ratio parameter of the training sample of complex insurance cases is identified by the neural network outputting the recent historical payout ratio parameter of similar cases. For the first Training labels for the actual payout ratio of a training sample of complex insurance cases. For the first The KL divergence between the insurance association weight distribution and the true weight training label distribution of a training sample of complex insurance cases. and The first The insurance association weight distribution and the true weight training label distribution of a complex insurance case training sample. For the first The total number of similar cases corresponding to the training sample of complex insurance cases. and The first The training sample for complex insurance cases is the same as the first one. The training sample corresponding to the complex insurance case The insurance association weights and true weights used to train the labels among similar cases described above. The parameter is the minimum non-zero parameter in the denominator.
8. An insurance case triage system, characterized in that, The system includes: The insurance case preliminary triage module is used to obtain the insurance case information of the insurance case and input the insurance case information into the insurance case triage rule engine to obtain the case type information and applicable insurance rules of the insurance case. The case type information includes ordinary case type information and compensation ratio restriction case type information. The complex insurance case identification module is used to identify insurance cases whose case type information is the compensation ratio restriction case type information as complex insurance cases, and to obtain the compensation ratio restriction rule and the compensation ratio parameter of the compensation ratio restriction rule in the applicable insurance rules of the complex insurance case; The recent historical case identification module is used to obtain recent historical cases that are within the rule engine update cycle and real-time case analysis cycle from the historical case library if the rule engine update interval of the insurance case diversion rule engine exceeds a preset rule engine degradation cycle threshold, and to construct a recent historical case library based on the recent historical cases. The similar insurance case identification module is used to identify similar cases of the complex insurance case from the recent historical case database according to the compensation ratio restriction rules of the complex insurance case; The payout ratio parameter calculation module is used to input the insurance case information of the complex insurance case, the insurance case information of each of the similar cases, and the actual payout ratio parameters of each of the similar cases into the payout ratio parameter recognition neural network to generate the recent historical similar payout ratio parameters of the complex insurance case corresponding to the payout ratio parameters; The claims settlement report generation module is used to generate an insurance claims settlement report for the complex insurance case based on the recent historical similar claims ratio parameters and the applicable insurance rules.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.