Insurance business process intelligent inspection system and method based on domain-specific language

Abstract

The invention relates to the technical field of computer software, and discloses an intelligent inspection system and method for an insurance business process based on a domain-specific language, and the system comprises a business process modeling and DSL module which carries out the unified abstraction and executable modeling of the insurance business process through the domain-specific language, and obtains a process model which can be executed by a machine; the intelligent inspection execution and data acquisition module drives the operation of the insurance business process through an execution process model, and acquires full-link business data of the insurance business process in the operation process; the service attribution analysis module is used for positioning an abnormal root cause in an insurance service process through an attribution algorithm based on the full-link service data; and the intelligent report generation module is used for generating a natural language inspection report through an AI large model based on the insurance business process context provided by the abnormal root cause and the process model. According to the invention, end-to-end monitoring, accurate root cause positioning and automatic intelligent report generation of the insurance service process can be realized, and the operation and maintenance efficiency of the insurance service is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer software, and in particular to an insurance business process intelligent inspection system and method based on a domain-specific language. BACKGROUND

[0002] With the deepening of the digital transformation of the insurance industry, business process intelligent inspection has become a key means to ensure operational quality. The existing technology mainly has four shortcomings when dealing with complex insurance business scenarios: First, the business process is fragmented, and each business system runs independently to form a "data island". There is a lack of unified monitoring perspective across the "quotation-insurance-claim" whole process, and end-to-end inspection is difficult to achieve. Second, the intelligent analysis is shallow, and the "black box" characteristics of AI models result in weak abnormal attribution ability and difficulty in locating bottleneck causes. Third, the system is rigid and lacks flexibility. Whether it is a closed system or a low-code platform, it is difficult to support business experts to directly and flexibly define and adjust complex inspection logic, and the response to business changes is slow. Fourth, the inspection results are not fully utilized, and the inspection data has not been deeply transformed into actionable insights. Analysis reports still rely on manual writing, and the value has not been maximized.

[0003] Therefore, there is an urgent need to provide a technical solution to solve the above problems. SUMMARY

[0004] To solve the above technical problems, the present application provides an insurance business process intelligent inspection system and method based on a domain-specific language.

[0005] In a first aspect, the present application provides an insurance business process intelligent inspection system based on a domain-specific language. The technical solution of the system is as follows: A business process modeling and DSL module is used to uniformly abstract and executable model the insurance business process through a domain-specific language, and obtain a machine executable process model; An intelligent inspection execution and data collection module is used to drive the operation of the insurance business process by executing the process model, and collect full-link business data of the insurance business process during operation; A business attribution analysis module is used to locate abnormal root causes in the insurance business process through attribution algorithms based on the full-link business data; An intelligent report generation module is used to generate natural language inspection reports through AI large models based on the abnormal root causes and the insurance business process context provided by the process model.

[0006] The beneficial effects of an insurance business process intelligent inspection system based on a domain-specific language of the present application are as follows: The system of this invention uses a domain-specific language to uniformly model and execute insurance business processes, combines attribution algorithms to locate the root causes of insurance business anomalies, and generates inspection reports based on the business process context using an AI large model. This solves the problems of data silos, shallow analysis, system rigidity, and reliance on manual reporting in insurance business systems, and realizes end-to-end monitoring of insurance business processes, accurate root cause location, and automated intelligent report generation, thereby improving the efficiency of insurance business operation and maintenance.

[0007] Based on the above solution, the intelligent inspection system for insurance business processes based on domain-specific languages ​​of the present invention can be further improved as follows.

[0008] In an alternative approach, the business process modeling and DSL module is specifically used for: Obtain the elements and logical relationships of the insurance business process defined using the domain-specific language; Parse the elements and logical relationships to obtain the business process components and their connections; Construct a unified business process abstract model based on the business process components and the connection relationships; The business process abstract model is subjected to semantic analysis and structural verification. The verified business process abstract model is then compiled to generate a machine-executable process model.

[0009] The advantages of adopting the above optional approach are: by further obtaining the elements and logical relationships defined by the DSL, and through the complete processing chain of parsing, construction, verification and compilation, the business process abstract model is transformed into a machine-executable process model, thereby improving the accuracy and executability of the model.

[0010] In one alternative approach, the intelligent inspection execution and data acquisition module is specifically used for: According to the execution logic defined in the process model, the quotation system, underwriting system and claims system that constitute the insurance business process are driven to perform corresponding business operations. During the process of driving the quotation system, the underwriting system and the claims system to perform business operations, the business status and performance indicators of the quotation system, the underwriting system and the claims system are recorded and integrated into the full-link business data.

[0011] The advantages of adopting the above optional approach are: further, according to the execution logic defined by the process model, it directly drives the three core systems of quotation, underwriting, and claims to perform business operations, and records the business status and performance indicators of each system, thereby achieving unified collection of business data across the entire chain.

[0012] In one alternative approach, the business attribution analysis module is specifically used for: Perform anomaly detection on the entire chain of business data to identify at least one abnormal node in the insurance business process; Based on the node dependencies defined in the process model, construct a causal graph of the insurance business process; Based on the cause-effect graph, backtracking analysis is performed on each abnormal node to determine the critical path corresponding to each abnormal node. The attribution algorithm calculates the influence weight of all nodes on the abnormal results in each critical path, and locates the node with the largest influence weight in each critical path as the corresponding root cause of the anomaly.

[0013] The advantages of adopting the above optional methods are: further anomaly detection and identification of abnormal nodes in the entire business data, construction of cause-effect graphs based on process models and backtracking analysis of critical paths, calculation of node influence weights through attribution algorithms, and accurate location of the root cause of anomalies.

[0014] In one alternative approach, the intelligent report generation module is specifically used for: Extract the insurance business process context from the process model, and combine the insurance business process context with the root cause of the anomaly corresponding to each critical path into structured input data; The structured input data is input into the AI ​​model to generate the natural language inspection report.

[0015] The advantages of adopting the above optional approach are: further extracting the business process context from the process model, combining the context with the root causes of anomalies in each critical path into structured input data, providing a high-quality information source for the AI ​​large model, and supporting the generation of natural language inspection reports.

[0016] In one alternative approach, the AI ​​big model is a pre-trained big language model, which is specifically used to generate the natural language inspection report containing root cause analysis and optimization suggestions by analyzing the correlation between business process context and anomaly root causes in the structured input data.

[0017] The benefits of adopting the above optional approach are: further using pre-trained large language models to analyze the correlation between business process context and root cause of anomalies in structured input data, automatically generating natural language inspection reports containing root cause analysis and optimization suggestions, and improving the report's insight.

[0018] In one alternative approach, the domain-specific language includes: underwriting rules for defining underwriting conditions for insurance products, rate calculation rules for defining premium calculation logic, claims rules for defining claims review conditions, and process connection rules for defining the data flow relationship between the quotation system, underwriting system, and claims system.

[0019] The beneficial effects of adopting the above optional methods are: further enabling domain-specific languages ​​to fully cover the entire business process of insurance product underwriting, premium calculation, claims review and data flow between systems through underwriting rules, premium calculation rules, claims rules and process connection rules.

[0020] Secondly, this invention provides an intelligent inspection method for insurance business processes based on a domain-specific language. The technical solution of this method is as follows: By using a domain-specific language to uniformly abstract and executablely model insurance business processes, a machine-executable process model is obtained. The insurance business process is driven by executing the process model, and full-link business data of the insurance business process is collected during the process. Based on the full-link business data, the root causes of anomalies in the insurance business process are located using attribution algorithms. Based on the root cause of the anomaly and the insurance business process context provided by the process model, a natural language inspection report is generated through an AI big data model.

[0021] The beneficial effects of the intelligent inspection method for insurance business processes based on domain-specific languages ​​of the present invention are as follows: The method of this invention uses a domain-specific language to uniformly model and execute insurance business processes, combines attribution algorithms to locate the root causes of insurance business anomalies, and generates inspection reports based on the business process context using an AI large model. This solves the problems of data silos, shallow analysis, system rigidity, and reliance on manual reporting in insurance business systems, and realizes end-to-end monitoring of insurance business processes, accurate root cause location, and automated intelligent report generation, thereby improving the efficiency of insurance business operation and maintenance.

[0022] Thirdly, the technical solution of an electronic device according to the present invention is as follows: It includes a memory, a processor, and a program stored in the memory and running on the processor, wherein the processor executes the program to implement the steps of the domain-specific language-based intelligent inspection method for insurance business processes as described in this invention.

[0023] Fourthly, the technical solution of a computer-readable storage medium provided by the present invention is as follows: The computer-readable storage medium stores instructions that, when read, cause the computer-readable storage medium to perform the steps of the domain-specific language-based intelligent inspection method for insurance business processes of the present invention.

[0024] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0025] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic diagram of an embodiment of an intelligent inspection system for insurance business processes based on a domain-specific language, according to the present invention. Figure 2 This is an overall architecture diagram of an embodiment of an intelligent inspection system for insurance business processes based on a domain-specific language, according to the present invention. Figure 3 This is a flowchart illustrating an embodiment of an intelligent inspection method for insurance business processes based on a domain-specific language according to the present invention. Figure 4 This is a schematic diagram of an embodiment of an electronic device according to the present invention. Detailed Implementation

[0026] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0027] Figure 1 This diagram illustrates the structure of an embodiment of an intelligent inspection system 100 for insurance business processes based on a domain-specific language, provided by the present invention. Figure 1 As shown, the domain-specific language-based intelligent inspection system 100 for insurance business processes includes: The Business Process Modeling and DSL Module 101 is used to perform unified abstraction and executable modeling of insurance business processes using a domain-specific language, resulting in a machine-executable process model.

[0028] Domain-specific languages ​​refer to specialized computer languages ​​designed for the insurance business domain, used to express business rules and process logic; for example, business personnel might use this language to write rules such as "If a building is over 30 years old, a fire risk assessment must be added." Insurance business processes refer to the continuous sequence of business steps in insurance operations, from quotation to underwriting to claims settlement; for example, when a business owner insures a factory building for property insurance, the system sequentially executes asset valuation, risk assessment, and policy generation. Process models refer to executable computational models generated after abstracting and describing insurance business processes using domain-specific languages; for example, transforming asset registration, risk assessment, premium calculation, and their relationships into a computer-executable flowchart structure.

[0029] The intelligent inspection execution and data acquisition module 102 is used to drive the operation of the insurance business process by executing the process model, and to collect the full-link business data of the insurance business process during the operation.

[0030] Among them, full-link business data refers to the collection of status and performance data generated at each stage of the insurance business process; for example, it includes data such as the calculation time of the valuation system, the risk rating of the underwriting system, and the loss assessment of the claims system.

[0031] The business attribution analysis module 103 is used to locate the root causes of anomalies in the insurance business process based on the full-link business data and through attribution algorithms.

[0032] Attribution algorithms refer to calculation methods used to analyze the causes of anomalies in business processes; for example, the Shapley value algorithm is used to calculate the impact of each node in the business process on claims delays. Root causes refer to the nodes that are the fundamental causes of problems in the insurance business process; for example, the root cause of claims delays may be the lack of expert on-site investigation in the disaster loss assessment stage.

[0033] The intelligent report generation module 104 is used to generate a natural language inspection report based on the root cause of the anomaly and the insurance business process context provided by the process model, using an AI big data model.

[0034] The insurance business process context refers to the business process relationships and rule information provided by the process model; for example, it includes related information such as the asset valuation results required for risk assessment and the policy validity verification required for claims processing. The AI ​​large-scale model refers to a large-scale artificial intelligence model with natural language understanding and generation capabilities; for example, using the GPT-4 model to analyze property insurance business data and generate inspection reports. The natural language inspection report refers to an easily understandable business analysis document generated by the AI ​​large-scale model; for example, a Chinese description such as, "This month's delays in corporate property insurance claims are mainly due to a 40% increase in the average time spent on disaster loss assessment."

[0035] The technical solution of this embodiment uses a domain-specific language to uniformly model and execute insurance business processes, combines attribution algorithms to locate the root causes of insurance business anomalies, and generates inspection reports based on the business process context through an AI large model. This solves the problems of data silos, shallow analysis, system rigidity, and reliance on manual reporting in insurance business systems, and realizes end-to-end monitoring of insurance business processes, accurate root cause location, and automated intelligent report generation, thereby improving the efficiency of insurance business operation and maintenance.

[0036] In an alternative embodiment, the business process modeling and DSL module 101 is specifically used for: Obtain the elements and logical relationships of the insurance business process defined using the domain-specific language.

[0037] Here, "element" refers to the basic building blocks of the insurance business process; for example, the asset registration stage, risk assessment stage, and policy generation stage are all elements of the business process. "Logical relationship" refers to the connections and dependencies between the various elements in the business process; for example, the risk assessment stage can only begin after the asset valuation stage is completed.

[0038] Specifically, the insurance business process definition is received from business experts through a domain-specific language editor. This definition includes the elements of the insurance business process, such as business activity nodes and decision nodes, as well as the logical relationships between the elements, such as sequential connection relationships, conditional dependency relationships, and data transfer relationships. Among them, business activity nodes include quotation activities, underwriting activities, and claims activities; decision nodes include underwriting condition judgment and rate calculation rule judgment; and logical relationships define the execution order and data flow path between business activity nodes.

[0039] The elements and logical relationships are analyzed to obtain the business process components and their connections.

[0040] In this context, a business process component refers to a business processing unit with a specific function; for example, an asset valuation model, a fire risk assessment engine, and a claims calculator are all business process components. A connection relationship refers to the data transfer and control flow relationships between business process components; for example, the asset valuation component transmits its assessment results to the risk assessment component.

[0041] Specifically, a domain-specific language parser is used to perform syntactic analysis on insurance business process elements, identifying three types of business activity nodes: quoting, underwriting, and claims, as well as two types of decision nodes: underwriting condition judgment and rate calculation rule judgment. Simultaneously, semantic analysis is performed on logical relationships to determine the execution order between business activity nodes through sequential connections, the conditional branching logic between decision nodes through conditional dependencies, and the data flow path between nodes through data transfer relationships. Based on the parsing results, business activity nodes are mapped to corresponding business process components, including quoting, underwriting, and claims components. Connection relationships, including execution order relationships and data transfer relationships between components, are extracted from the logical relationships.

[0042] A unified business process abstract model is constructed based on the business process components and the connection relationships.

[0043] Among them, the business process abstract model refers to the structured description of the insurance business process; for example, a directed graph is used to represent the sequential relationship between asset registration, risk assessment, premium calculation and other steps.

[0044] Specifically, the quotation component, underwriting component, and claims component in the business process, as well as the underwriting condition judgment and rate calculation rule judgment in the decision node, are taken as the activity nodes of the abstract model, and the sequential connection relationship, conditional dependency relationship, and data transmission relationship in the connection relationship are taken as the edges of the abstract model. By combining these activity nodes and edges, a directed graph structure is formed. This directed graph structure fully represents the execution order, conditional branch logic, and data flow path of the insurance business process, thereby constructing a unified business process abstract model.

[0045] The business process abstract model is subjected to semantic analysis and structural verification. The verified business process abstract model is then compiled to generate a machine-executable process model.

[0046] Semantic analysis refers to checking whether the business process definition conforms to business logic; for example, verifying whether the risk assessment stage has configured the necessary disaster risk check rules. Structural verification refers to checking the structural integrity of the business process model; for example, ensuring that each property insurance business process has clearly defined application and policy generation nodes.

[0047] Specifically, semantic analysis is used to check whether the business rules and logical relationships in the business process abstract model meet the needs of insurance business. This includes verifying whether the underwriting conditions are consistent with the underwriting conditions of the insurance product and whether the rate calculation rules match the premium calculation logic. Structural verification is used to check the structural integrity of the business process abstract model, including confirming that all business activity nodes and decision nodes have valid connections and that there are no isolated nodes or circular dependencies. The business process abstract model that has undergone semantic analysis and structural verification is then input into the compiler. The compiler converts the nodes and edges in the abstract model into executable code, generating a machine-executable process model that can be directly loaded and executed by the inspection process engine.

[0048] In the above optional methods, the business process abstract model is further transformed into a machine-executable process model by obtaining the elements and logical relationships defined by the DSL and going through the complete processing chain of parsing, construction, verification and compilation, thereby improving the accuracy and executability of the model.

[0049] In one alternative embodiment, the intelligent inspection execution and data acquisition module 102 is specifically used for: According to the execution logic defined in the process model, the quotation system, underwriting system and claims system that constitute the insurance business process are driven to perform corresponding business operations.

[0050] The execution logic refers to the order in which business processes are executed as defined in the process model; for example, asset registration is performed first, followed by risk assessment, and finally premium calculation. The quotation system refers to the business system that processes insurance product quotation calculations; for example, a computer system that receives factory asset information and returns the premium amount. The underwriting system refers to the business system that processes insurance application reviews; for example, a computer system that assesses factory fire risks and decides whether to underwrite. The claims system refers to the business system that processes insurance claim applications; for example, a computer system that reviews disaster loss materials and calculates compensation amounts. Business operations refer to the specific actions performed in the insurance business process; for example, asset valuation, risk assessment, and claims settlement are all business operations.

[0051] Specifically, according to the execution logic defined in the process model, the machine-executable process model is loaded, and the node sequence and condition branches in the model are parsed. Based on the parsing results, the business interface of the quotation system is called in sequence to perform premium calculation operations, the business interface of the underwriting system is called to perform risk assessment and underwriting operations, and the business interface of the claims system is called to perform claims review and payment operations. During the calling process, the correct execution order and data transmission of each system operation are ensured based on the sequential connection relationship and condition dependency relationship in the process model.

[0052] During the process of driving the quotation system, the underwriting system and the claims system to perform business operations, the business status and performance indicators of the quotation system, the underwriting system and the claims system are recorded and integrated into the full-link business data.

[0053] Here, "business status" refers to the outcome of a business operation; for example, valuation completion, risk approval, and claims settlement are all business statuses. "Performance metrics" refers to data that measures the efficiency of business operation execution; for example, asset valuation taking 300 milliseconds or risk assessment accuracy of 95%.

[0054] Specifically, during the execution of business operations by the quotation system, underwriting system, and claims system, a unified data collector captures in real time the premium calculation status and response time of the quotation system, the underwriting results and risk assessment time of the underwriting system, and the claims review status and case processing time of the claims system. The captured business status and performance indicators are then cleaned and standardized in format, and aligned according to timestamps and business process instance identifiers, ultimately integrating them to form full-link business data containing complete business link status information and performance indicators.

[0055] In the above-mentioned optional methods, the execution logic defined by the process model is further used to directly drive the three core systems of quotation, underwriting, and claims to perform business operations, and record the business status and performance indicators of each system to achieve unified collection of business data across the entire chain.

[0056] In an alternative embodiment, the business attribution analysis module 103 is specifically used for: Anomaly detection is performed on the entire business data to identify at least one abnormal node in the insurance business process.

[0057] Anomaly detection refers to the process of identifying abnormal situations in the business process; for example, discovering that the average time for loss calculation in the claims process exceeds the threshold of 2 hours. Anomaly nodes refer to the specific steps in the business process where problems occur; for example, the underwriting process is suspended due to a lack of on-site inspection photos.

[0058] Specifically, business status and performance indicators are checked based on preset thresholds or statistical analysis methods. When performance indicators such as response time exceed the set threshold or business status such as transaction failure rate deviates from the normal range, the corresponding business process is marked as abnormal. Then, based on the mapping relationship between the marked abnormal indicators and insurance business process nodes, at least one abnormal node is identified. For example, the quotation node is identified as abnormal due to excessive response time or the underwriting node is identified as abnormal due to risk assessment failure.

[0059] Based on the node dependencies defined in the process model, a causal graph of the insurance business process is constructed.

[0060] Node dependency refers to the sequential dependence of various nodes in a business process; for example, the claims node depends on the result of the policy validity verification node. A cause-effect graph is a graph structure that represents the causal relationships between nodes in a business process; for example, using directed edges to connect asset registration, risk assessment, and claims processing nodes to represent dependencies.

[0061] Specifically, node dependencies are extracted from the process model, including sequential connection relationships that define the execution order of business activity nodes and conditional dependency relationships that define the branching logic of decision nodes; these node dependencies are mapped to edges of a directed graph, and business activity nodes and decision nodes are mapped to vertices of a directed graph; a causal graph is constructed based on the mapping results, which fully represents the causal relationships and data dependency paths between nodes in the insurance business process.

[0062] Based on the cause-effect graph, backtracking analysis is performed on each abnormal node to determine the critical path corresponding to each abnormal node.

[0063] Backtracking analysis refers to tracing back from anomaly points to possible causes; for example, tracing back from claims delays to incomplete disaster loss assessment data. The critical path refers to the sequence of nodes in a business process that has the greatest impact on overall execution; for example, asset registration → risk assessment → premium calculation constitute the critical path.

[0064] Specifically, starting from each abnormal node, the system traverses in the reverse direction of the node dependencies in the causal graph to collect all path sequences from the starting node of the business process to the abnormal node. Then, based on the business importance of the nodes in the path sequence, the depth of the dependency relationship, and historical abnormal data, the system evaluates the impact of each path on the abnormal result and determines the path with the greatest impact as the critical path corresponding to the abnormal node.

[0065] The attribution algorithm calculates the influence weight of all nodes on the abnormal results in each critical path, and locates the node with the largest influence weight in each critical path as the corresponding root cause of the anomaly.

[0066] The influence weight refers to the quantitative value of the degree of influence of a node on business process anomalies; for example, the influence weight of a risk assessment node on overall anomalies is 0.8.

[0067] Specifically, for each critical path, an attribution algorithm is used to quantify the contribution of each node to the abnormal result based on node performance indicators and business status data. The attribution algorithm uses the Shapley value method to evaluate the marginal impact of nodes in the path. Then, the influence weight values ​​of all nodes in the critical path are compared, and the node with the largest influence weight value is identified and located as the root cause of the anomaly in the corresponding critical path.

[0068] In the above-mentioned optional methods, anomaly detection is further performed on the entire business data to identify abnormal nodes, a cause-effect graph is constructed based on the process model and the critical path is backtracked and analyzed, and the influence weight of nodes is calculated through attribution algorithms to achieve accurate location of the root cause of the anomaly.

[0069] In one alternative embodiment, the intelligent report generation module 104 is specifically used for: The insurance business process context is extracted from the process model, and the insurance business process context is combined with the root cause of the anomaly corresponding to each critical path to form structured input data.

[0070] Structured input data refers to business data organized in a fixed format; for example, JSON data containing timestamps, asset types, risk levels, and claims status.

[0071] Specifically, extracting the insurance business process context from the process model includes obtaining the node attributes, node dependencies, and business rule information defined in the process model; combining the insurance business process context with the root causes of anomalies corresponding to each critical path into structured input data includes organizing the context information and root cause information according to a predetermined data structure, wherein the context information includes the business process name, dependency configuration, and business rule content, and the root cause information includes the critical path identifier, the type of abnormal node, and performance indicator data, forming machine-readable structured input data.

[0072] The structured input data is input into the AI ​​model to generate the natural language inspection report.

[0073] Specifically, the process of inputting structured input data into an AI big model to generate a natural language inspection report includes: sending structured input data to a pre-trained big language model via the AI ​​big model interface; the pre-trained big language model parsing the insurance business process context and anomaly root cause information in the structured input data, and analyzing the correlation between the business process context and the anomaly root cause; based on the correlation analysis results, generating natural language text containing root cause analysis and optimization suggestions; and the report renderer receiving the natural language text and formatting it into a standardized natural language inspection report.

[0074] In the above-mentioned optional methods, the business process context is further extracted from the process model, and the context is combined with the root cause of the anomaly of each critical path into structured input data to provide a high-quality information source for the AI ​​big model and support the generation of natural language inspection reports.

[0075] In one alternative approach, the AI ​​big model is a pre-trained big language model, which is specifically used to generate the natural language inspection report containing root cause analysis and optimization suggestions by analyzing the correlation between business process context and anomaly root causes in the structured input data.

[0076] Among these, pre-trained large language models refer to language generation models trained on massive amounts of text; for example, the GPT-4 model trained on trillions of texts. Root cause analysis refers to exploring the underlying causes of anomalies; for example, analyzing and concluding that claims delays are due to a lack of professional engineering personnel in disaster loss assessment. Optimization suggestions refer to improvement plans proposed for the root causes of anomalies; for example, suggesting the participation of third-party assessment agencies in major disaster loss assessments.

[0077] Among the above optional methods, a pre-trained large language model is further used to analyze the correlation between business process context and root cause of anomalies in structured input data, and to automatically generate a natural language inspection report containing root cause analysis and optimization suggestions, thereby improving the report's insight.

[0078] In one alternative approach, the domain-specific language includes: underwriting rules for defining underwriting conditions for insurance products, rate calculation rules for defining premium calculation logic, claims rules for defining claims review conditions, and process connection rules for defining the data flow relationship between the quotation system, underwriting system, and claims system.

[0079] Among these, underwriting conditions refer to the conditions that must be met before an insurance product can be underwritten; for example, corporate property insurance requires buildings to meet the latest fire safety standards. Underwriting rules refer to the business rules used in the underwriting process; for example, "reinsurance arrangements are required for insured assets exceeding 50 million" is an underwriting rule. Premium calculation logic refers to the business rules for calculating insurance prices; for example, property insurance premium = insured amount × rate × discount factor. Rate calculation rules refer to the methods for determining insurance rates; for example, different rate tables are determined based on building structure type and usage. Claims review conditions refer to the conditions that must be met for claims processing; for example, property insurance claims require proof of disaster loss and a repair quote. Claims rules refer to the business rules used in the claims process; for example, "losses caused by natural disasters are compensated at 80% of the actual repair costs" is a claims rule. Data flow relationships refer to the data transmission relationships between systems; for example, the quotation system transmits the policy number to the claims system. Process connection rules refer to the rules that define the connection relationships between business process steps; for example, the premium calculation process is automatically triggered after the risk assessment is passed.

[0080] Among the above-mentioned optional methods, further underwriting rules, premium calculation rules, claims rules, and process connection rules are used to ensure that domain-specific languages ​​fully cover the entire business process of insurance product underwriting, premium calculation, claims review, and data flow between systems.

[0081] In this embodiment, as Figure 2 As shown: 1) The Business Process Modeling and DSL Module 101 includes: a DSL editor and visual interface, and a DSL parser. The DSL editor and visual interface receive insurance business process elements and logical relationships defined in a domain-specific language. The DSL parser parses the elements and logical relationships to obtain business process components and connections. Based on the business process components and connections, a unified business process abstract model is constructed. After semantic analysis and structural verification of the business process abstract model, the verified business process abstract model is compiled to generate a machine-executable process model.

[0082] 2) The intelligent inspection execution and data acquisition module 102 includes an inspection process engine and a unified data collector. The inspection process engine drives the operation of the insurance business process through the execution process model, while the unified data collector collects the full-link business data of the insurance business process during the operation.

[0083] 3) The business attribution analysis module 103 includes an attribution algorithm library and a root fixer. The attribution algorithm library performs anomaly detection and identifies abnormal nodes in the insurance business process based on full-link business data. The root fixer constructs a cause-effect graph of the insurance business process according to the node dependency relationship defined in the process model. Based on the cause-effect graph, it performs backtracking analysis on each abnormal node to determine the critical path corresponding to each abnormal node. Through the attribution algorithm, it calculates the influence weight of all nodes on the abnormal result in each critical path and locates the node with the largest influence weight in each critical path as the corresponding abnormal root cause.

[0084] 4) The intelligent report generation module 104 includes an AI big model interface and a report renderer. The AI ​​big model interface extracts the insurance business process context from the process model and combines the insurance business process context with the root cause of the anomaly corresponding to each critical path into structured input data. The structured input data is input into the AI ​​big model. The report renderer receives the natural language content generated by the AI ​​big model and outputs a natural language inspection report.

[0085] To better illustrate the technical solution of this embodiment, the following example is used for complete explanation: S10. The domain-specific language editor receives the enterprise property insurance business process definition input by business experts. The definition includes three business activity nodes: asset valuation activity, risk assessment activity, and policy generation activity; two decision nodes: building age judgment and fire risk level judgment; and three logical relationships: the sequential connection relationship defining the execution order of each node, the conditional dependency relationship defining the conditional branch, and the data transfer relationship defining the data transfer. S20, Business Process Modeling and DSL Module 101 acquires insurance business process elements and logical relationships defined using a domain-specific language, parses out three business process components—asset valuation component, risk assessment component, and policy generation component—as well as the connection relationships between components, constructs a unified business process abstract model based on these components and connection relationships, performs semantic analysis and structural verification on the abstract model, and compiles the verified business process abstract model into a machine-executable process model. S30, Intelligent Inspection Execution and Data Acquisition Module 102, according to the execution logic defined in the process model, sequentially drives the asset valuation system to perform asset value assessment, drives the risk assessment system to perform fire risk investigation, and drives the policy system to perform policy generation. In the process of driving each system to perform business operations, it records the assessment status and calculation time of the asset valuation system, the risk level and investigation time of the risk assessment system, and the generation status and processing time of the policy system, and integrates these business statuses and performance indicators into full-link business data. S40, Business Attribution Analysis Module 103 uses full-link business data to detect anomalies and identify risk assessment nodes as anomalies due to timeouts in on-site investigations. Based on the node dependencies defined in the process model, a causal graph containing asset valuation → risk assessment → policy generation is constructed. Based on the causal graph, a backtracking analysis of the risk assessment anomaly nodes is performed to determine that the path is a critical path. The attribution algorithm is used to calculate the influence weight of all nodes on the abnormal results in the critical path and locates the risk assessment node as the root cause of the anomaly. S50 and the intelligent report generation module 104 extract the insurance business process context, which includes the business process dependencies and underwriting rules, from the process model. They combine the business process context with the root causes of risk assessment anomalies corresponding to the critical path into structured input data. The structured input data is then input into a pre-trained large language model to generate a natural language inspection report containing root cause analysis and optimization suggestions. The report points out that the fire risk assessment process is delayed due to insufficient on-site investigation resources and recommends increasing the number of investigation personnel.

[0086] Figure 3This diagram illustrates an embodiment of an intelligent inspection method for insurance business processes based on a domain-specific language, provided by the present invention. This method can be executed by electronic devices such as terminal devices or servers. The terminal device can be any fixed or mobile terminal, such as user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (PDA), handheld device, computing device, in-vehicle device, or wearable device. The server can be a single server or a server cluster consisting of multiple servers. Any electronic device can implement the intelligent inspection method for insurance business processes based on a domain-specific language by having its processor call computer-readable instructions stored in its memory. Figure 3 As shown, it includes the following steps: S1. By using a domain-specific language to uniformly abstract and executablely model the insurance business process, a machine-executable process model is obtained. S2. Drive the operation of the insurance business process by executing the process model, and collect full-link business data of the insurance business process during the operation; S3. Based on the full-link business data, locate the root cause of anomalies in the insurance business process using an attribution algorithm; S4. Based on the root cause of the anomaly and the insurance business process context provided by the process model, a natural language inspection report is generated through the AI ​​big data model.

[0087] In one alternative approach, S1 specifically includes: Obtain the elements and logical relationships of the insurance business process defined using the domain-specific language; Parse the elements and logical relationships to obtain the business process components and their connections; Construct a unified business process abstract model based on the business process components and the connection relationships; The business process abstract model is subjected to semantic analysis and structural verification. The verified business process abstract model is then compiled to generate a machine-executable process model.

[0088] In one alternative approach, S2 specifically includes: According to the execution logic defined in the process model, the quotation system, underwriting system and claims system that constitute the insurance business process are driven to perform corresponding business operations. During the process of driving the quotation system, the underwriting system and the claims system to perform business operations, the business status and performance indicators of the quotation system, the underwriting system and the claims system are recorded and integrated into the full-link business data.

[0089] In one alternative approach, S3 specifically includes: Perform anomaly detection on the entire chain of business data to identify at least one abnormal node in the insurance business process; Based on the node dependencies defined in the process model, construct a causal graph of the insurance business process; Based on the cause-effect graph, backtracking analysis is performed on each abnormal node to determine the critical path corresponding to each abnormal node. The attribution algorithm calculates the influence weight of all nodes on the abnormal results in each critical path, and locates the node with the largest influence weight in each critical path as the corresponding root cause of the anomaly.

[0090] In one alternative approach, S4 specifically includes: Extract the insurance business process context from the process model, and combine the insurance business process context with the root cause of the anomaly corresponding to each critical path into structured input data; The structured input data is input into the AI ​​model to generate the natural language inspection report.

[0091] In one alternative approach, the AI ​​big model is a pre-trained big language model, which is specifically used to generate the natural language inspection report containing root cause analysis and optimization suggestions by analyzing the correlation between business process context and anomaly root causes in the structured input data.

[0092] In one alternative approach, the domain-specific language includes: underwriting rules for defining underwriting conditions for insurance products, rate calculation rules for defining premium calculation logic, claims rules for defining claims review conditions, and process connection rules for defining the data flow relationship between the quotation system, underwriting system, and claims system.

[0093] It should be noted that the beneficial effects of the domain-specific language-based intelligent inspection method for insurance business processes provided in the above embodiments are the same as those of the domain-specific language-based intelligent inspection system 100 for insurance business processes, and will not be repeated here. Furthermore, the system provided in the above embodiments is only illustrated by the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the system can be divided into different functional modules according to the actual situation to complete all or part of the functions described above. In addition, the method and system embodiments provided in the above embodiments belong to the same concept, and their specific implementation process is detailed in the system embodiments, and will not be repeated here.

[0094] The domain-specific language-based intelligent inspection system 100 of the present invention can be a computer program (including program code) running on a computer device. For example, the domain-specific language-based intelligent inspection system 100 of the present invention is an application software that can be used to execute the corresponding steps in the domain-specific language-based intelligent inspection method of the present invention.

[0095] In some embodiments, the domain-specific language-based intelligent inspection system 100 for insurance business processes of the present invention can be implemented using a combination of hardware and software. As an example, the domain-specific language-based intelligent inspection system 100 for insurance business processes of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the domain-specific language-based intelligent inspection method for insurance business processes of the present invention. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0096] The modules described in the embodiments of this invention can be implemented in software or hardware. The names of the modules are not, in some cases, limiting the scope of the module itself.

[0097] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-mentioned intelligent inspection methods for insurance business processes based on a domain-specific language. That is, an electronic device according to an embodiment of the present invention may include, but is not limited to: a processor and a memory; the memory is used to store the computer program; the processor is used to execute the intelligent inspection method for insurance business processes based on a domain-specific language as shown in any embodiment of the present invention by calling the computer program.

[0098] In one alternative embodiment, an electronic device is provided, such as Figure 4 As shown, Figure 4The illustrated electronic device 4000 includes a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, for example, via a bus 4002. Optionally, the electronic device 4000 may further include a transceiver 4004, which can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 4004 is not limited to one type, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present invention.

[0099] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 4001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0100] Bus 4002 may include a path for transmitting information between the aforementioned components. Bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 4002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The bus 4002 is represented by only one thick line, but this does not mean that there is only one bus or one type of bus.

[0101] The memory 4003 may be ROM (Read Only Memory) or other types of static storage devices capable of storing static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices capable of storing information and instructions, or EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0102] The memory 4003 stores application code (computer program) for executing the present invention, and its execution is controlled by the processor 4001. The processor 4001 executes the application code stored in the memory 4003 to implement the content shown in the foregoing method embodiments.

[0103] Among them, electronic devices can also be terminal devices. A terminal device can be any terminal device that can install applications and access web pages through applications, including at least one of smartphones, tablets, laptops, desktop computers, smart speakers, smartwatches, smart TVs, and smart in-vehicle devices.

[0104] It should be noted that, Figure 4 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0105] An embodiment of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-mentioned intelligent inspection methods for insurance business processes based on a domain-specific language.

[0106] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0107] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. The processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the aforementioned domain-specific language-based intelligent inspection method for insurance business processes.

[0108] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0109] It should be understood that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0110] The computer-readable storage medium provided in this invention can be, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0111] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the method shown in the above embodiments.

[0112] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0113] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0114] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this invention can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0115] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. An intelligent inspection system for insurance business processes based on a domain-specific language, characterized in that, include: The Business Process Modeling and DSL module is used to perform unified abstraction and executable modeling of insurance business processes using a domain-specific language, resulting in a machine-executable process model. The intelligent inspection execution and data acquisition module is used to drive the operation of the insurance business process by executing the process model, and to collect the full-link business data of the insurance business process during the operation. The business attribution analysis module is used to locate the root causes of anomalies in the insurance business process based on the full-link business data and through attribution algorithms. The intelligent report generation module is used to generate natural language inspection reports based on the root cause of the anomaly and the insurance business process context provided by the process model, using an AI big data model.

2. The intelligent inspection system for insurance business processes based on domain-specific languages ​​according to claim 1, characterized in that, The business process modeling and DSL module is specifically used for: Obtain the elements and logical relationships of the insurance business process defined using the domain-specific language; Parse the elements and logical relationships to obtain the business process components and their connections; Construct a unified business process abstract model based on the business process components and the connection relationships; The business process abstract model is subjected to semantic analysis and structural verification. The verified business process abstract model is then compiled to generate a machine-executable process model.

3. The intelligent inspection system for insurance business processes based on a domain-specific language according to claim 2, characterized in that, The intelligent inspection execution and data acquisition module is specifically used for: According to the execution logic defined in the process model, the quotation system, underwriting system and claims system that constitute the insurance business process are driven to perform corresponding business operations. During the process of driving the quotation system, the underwriting system and the claims system to perform business operations, the business status and performance indicators of the quotation system, the underwriting system and the claims system are recorded and integrated into the full-link business data.

4. The intelligent inspection system for insurance business processes based on domain-specific languages ​​according to claim 1, characterized in that, The business attribution analysis module is specifically used for: Perform anomaly detection on the entire chain of business data to identify at least one abnormal node in the insurance business process; Based on the node dependencies defined in the process model, construct a causal graph of the insurance business process; Based on the cause-effect graph, backtracking analysis is performed on each abnormal node to determine the critical path corresponding to each abnormal node. The attribution algorithm calculates the influence weight of all nodes on the abnormal results in each critical path, and locates the node with the largest influence weight in each critical path as the corresponding root cause of the anomaly.

5. The intelligent inspection system for insurance business processes based on a domain-specific language according to claim 4, characterized in that, The intelligent report generation module is specifically used for: Extract the insurance business process context from the process model, and combine the insurance business process context with the root cause of the anomaly corresponding to each critical path into structured input data; The structured input data is input into the AI ​​model to generate the natural language inspection report.

6. The intelligent inspection system for insurance business processes based on domain-specific languages ​​according to claim 5, characterized in that, The AI ​​big model is a pre-trained big language model. Specifically, the AI ​​big model is used to generate a natural language inspection report containing root cause analysis and optimization suggestions by analyzing the correlation between business process context and anomaly root causes in the structured input data.

7. The intelligent inspection system for insurance business processes based on domain-specific language according to claim 3, characterized in that, The domain-specific languages ​​include: underwriting rules for defining the underwriting conditions of insurance products, rate calculation rules for defining the premium calculation logic, claims rules for defining the claims review conditions, and process connection rules for defining the data flow relationship between the quotation system, the underwriting system, and the claims system.

8. A method for intelligent inspection of insurance business processes based on a domain-specific language, characterized in that, include: By using a domain-specific language to uniformly abstract and executablely model insurance business processes, a machine-executable process model is obtained. The insurance business process is driven by executing the process model, and full-link business data of the insurance business process is collected during the process. Based on the full-link business data, the root causes of anomalies in the insurance business process are located using attribution algorithms. Based on the root cause of the anomaly and the insurance business process context provided by the process model, a natural language inspection report is generated through an AI big data model.

9. An electronic device, characterized in that, The electronic device includes a processor coupled to a memory, the memory storing at least one computer program, which is loaded and executed by the processor to enable the electronic device to implement the domain-specific language-based intelligent inspection method for insurance business processes as described in claim 8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which, when executed by a processor, implements the intelligent inspection method for insurance business processes based on a domain-specific language as described in claim 8.

Images