Insurance information monitoring method and device, computer equipment and storage medium

Abstract

The invention discloses an insurance information monitoring method and device, computer equipment and a storage medium. The method comprises the following steps: establishing connection with a monitored insurance system through a preset monitoring system; obtaining an insurance data object to be monitored from the monitored insurance system; monitoring the to-be-monitored insurance data object through an insurance data monitoring model accumulated in a preset monitoring system to obtain monitoring operation data; and judging the operation condition of the monitored insurance system according to the monitoring operation data, and calling a processing mode corresponding to the operation condition according to an exception processing library in a preset monitoring system to solve the operation condition. In the scheme, by constructing the monitoring model library in the preset monitoring system, the management and reuse of the monitoring data and the model can be realized directly through the insurance data object and the model object; by constructing the exception handling library in the preset monitoring system, the exception problem can be directly matched through the exception handling mode.

Description

Technical Field

[0001] The present invention relates to the fields of monitoring and financial technology, and in particular to an insurance information monitoring method, device, computer equipment and storage medium. Background Art

[0002] The increasing informatization of the insurance industry has led to a corresponding increase in the number of insurance systems. The abnormal monitoring and alarm information generated during insurance system operation requires system operations and maintenance personnel to promptly locate anomalies and identify the operating patterns of the insurance systems. However, traditional insurance system operations and maintenance methods only detect failures when they occur or impact business continuity. This passive firefighting approach not only makes it impossible to accurately locate the anomaly due to delayed abnormal data collection, but more seriously, it can lead to a vicious chain reaction of failures. Furthermore, monitoring to ensure the stable operation of the insurance system and the normal flow of business requires a significant amount of manpower and material resources.

[0003] Although developers can automatically avoid risks for some common abnormal problems after writing system programs, some abnormalities still require system operation and maintenance personnel to participate in real-time monitoring and timely processing. For example, some parameters in the system operation will be dynamically adjusted according to the current operating status. Therefore, the current insurance system still requires a lot of manpower and material resources for monitoring.

[0004] Therefore, it is urgent to find a new technical solution to solve the above problems. Summary of the Invention

[0005] Based on this, an insurance information monitoring method, device, computer equipment and storage medium are provided to solve the technical problem in the prior art that the insurance system cannot be automatically monitored and requires a large amount of manpower and material resources for monitoring.

[0006] An insurance information monitoring method, comprising: Establishing a connection with the monitored insurance system through a preset monitoring system; Acquiring an insurance data object to be monitored from the monitored insurance system; Performing monitoring processing on the insurance data object to be monitored by using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; The operation status of the monitored insurance system is determined according to the monitoring operation data, and a processing method corresponding to the operation status is called according to the exception processing library in the preset monitoring system to solve the operation status.

[0007] An insurance information monitoring device, comprising: An establishment module for establishing a connection with a monitored insurance system through a preset monitoring system; An acquisition module, configured to acquire the insurance data object to be monitored from the monitored insurance system; A monitoring processing module, configured to perform monitoring processing on the insurance data object to be monitored using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; The judgment module is used to judge the operation status of the monitored insurance system according to the monitoring operation data, and call the processing method corresponding to the operation status according to the exception processing library in the preset monitoring system to solve the operation status.

[0008] A computer device 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, the insurance information monitoring method is implemented.

[0009] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the insurance information monitoring method is implemented.

[0010] The above-mentioned insurance information monitoring method, device, computer equipment and storage medium establish a connection with the monitored insurance system through a preset monitoring system; obtain the insurance data object to be monitored from the monitored insurance system; monitor and process the insurance data object to be monitored through the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; judge the operating status of the monitored insurance system based on the monitoring operation data, and call the processing method corresponding to the operating status according to the exception handling library in the preset monitoring system to solve the operating status. In this solution, by constructing a monitoring model library in a preset monitoring system, it is possible to directly use insurance data objects and model objects (specifically, the monitored insurance system can directly call insurance data objects and model objects), reduce the duplication of work caused by creating monitoring objects by modifying configuration files, and realize the management and reuse of monitoring data and models; by constructing an exception handling library in a preset monitoring system, it is possible to directly match exception problems through exception handling methods (specifically, the monitored insurance system can directly call exception handling methods), reduce the duplication of work to meet the same exception handling needs of different systems, realize the management and reuse of exception handling methods by the monitored insurance system, and improve the reusability and configurability of exception handling; in this way, both can reduce the configuration of manpower and material resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0012] Figure 1 This is a schematic diagram of an application environment of an insurance information monitoring method according to an embodiment of the present invention; Figure 2 is a process diagram of an insurance information monitoring method according to an embodiment of the present invention; Figure 3 It is a structural diagram of an insurance information monitoring device according to one embodiment of the present invention; Figure 4 FIG. 1 is a schematic diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0013] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0014] The insurance information monitoring method provided by the present invention can be applied in Figure 1 In an application environment, a client communicates with a server via a network. The client may include, but is not limited to, various personal computers, laptops, smartphones, tablet computers, and portable wearable devices. The server may be implemented as a standalone server or a server cluster consisting of multiple servers. For example, the present invention may be applied to the server corresponding to the data middleware.

[0015] In one embodiment, if Figure 2 As shown, a method for monitoring insurance information is provided, which is applied in Figure 1 The following steps are used as an example to illustrate the server where the platform is located: S10, establishing a connection with the monitored insurance system through a preset monitoring system.

[0016] Understandably, the preset monitoring system can be an automated system that is built, which may include a data sampling information library, a monitoring model library, an exception handling library, a network communication module, a data collection module, a data exception handling and analysis module, a page display module, and a monitoring database. Each module cooperates with each other and can realize corresponding functions; the monitored insurance system can be a system used by users, such as a policy correction system, a group quotation system, and a database cluster; the monitored insurance system opens the JMX access port and establishes a remote access connection between the preset monitoring system and the monitored insurance system. Specifically, it needs to be implemented step by step from four dimensions: parameter configuration, security reinforcement, network connection, and tool integration.

[0017] It should be noted that the pre-set monitoring system can be implemented based on JMX. JMX (Java Management Extensions) defines the architecture, design patterns, application program interfaces, and services for application and network management and monitoring in the Java programming language. Using JMX to monitor the system's operating status or manage certain aspects of the system, such as clearing caches and reloading configuration files, can easily enable applications to have a managed and scalable architecture. Each JMXAgent service can be easily integrated into the Agent.

[0018] S20: Obtain the insurance data object to be monitored from the monitored insurance system.

[0019] Understandably, the insurance data objects to be monitored may be data objects at various levels and dimensions, such as the fluctuation range of the regional risk coefficient and the quotation request QPS on the quotation system, the master-slave synchronization delay and the slow query ratio on the database cluster, and the maximum number of connections and the idle timeout on the policy modification system. Specifically, according to the types of different systems to be monitored, the insurance data objects to be monitored with higher priorities can be determined, or according to the types of different systems to be monitored, the data objects with the most historical selections can be determined as the insurance data objects to be monitored. Similarly, the user can also manually select data objects as the insurance data to be monitored.

[0020] S30: Monitoring and processing the insurance data object to be monitored is performed using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data.

[0021] It can be understood that the preset monitoring system in the preset monitoring system is a model accumulated after system training, which is ultimately presented as a collection model. Its normal workflow can be A (obtaining original indicators), B (feature calculation, obtaining target features) C (determine the baseline model), D (add the target features to the training set), E (determine the training results of the training set using the baseline model and determine whether the training results deviate from the baseline), F (multi-model voting, | confirm anomaly | G [generate an alarm] | suspected anomaly | H [manual annotation feedback]); monitoring operation data can refer to real-time or historical data collected, stored, and analyzed through technical means to evaluate the performance, health status, and resource usage of the system or application during operation. Examples include operational data related to the monitored insurance data objects in various types of systems, the fluctuation range of regional risk factors, and the QPS of quote requests.

[0022] S40, judging the operating status of the monitored insurance system according to the monitoring operating data, and calling a processing method corresponding to the operating status according to the exception processing library in the preset monitoring system to solve the operating status.

[0023] Understandably, the monitoring operation data can be verified against the threshold data in the preset monitoring system, and each verification result corresponds to the operation status. In this way, the operation status of the monitored insurance system can be determined by the verification result, such as whether the regional risk coefficient of the system fluctuates greatly or the size of the quotation request QPS is large. The processing method is to match the solution to the problem according to the operation status. For example, if the operation status exceeds the alarm threshold, an alarm will be processed or other corresponding restrictions will be processed. Taking steps S10 to S40 as an example, in the elastic management of the database connection pool of the insurance policy revision system, the insurance data objects to be monitored are the maximum number of connections and the idle timeout period. The business scenario can temporarily increase the maximum number of connections from 200 to 500 during the Double Eleven promotion. The monitoring implementation method is to monitor the Druid connection pool MXBean through JMX. The operation status is that the connection is not released for more than 300ms (setting the connection leak detection threshold), and the processing method is automatic alarm processing; in the group finance quotation system, the insurance data to be monitored are the fluctuation range of the regional risk coefficient and the quotation request QPS. The operation status is the daily change of the coefficient in a certain area. The processing method is to trigger an alarm at 15%. The operation status is that the QPS exceeds 5000 times / second, and the processing method is automatic current limiting; in the database cluster, the insurance data to be monitored are the master-slave synchronization delay and the slow query ratio. The operation status is the synchronization delay of 500ms. The processing method is to trigger the master-slave switch. The operation status is that the slow query ratio is 5%, and the processing method is to automatically initiate index optimization. In the embodiment of steps S10 to S40, by constructing a monitoring model library in a preset monitoring system, the repeated work caused by creating monitoring objects by modifying configuration files can be reduced directly through insurance data objects and model objects (specifically, the monitored insurance system can directly call insurance data objects and model objects), thereby realizing the management and reuse of monitoring data and models; by constructing an exception handling library in a preset monitoring system, the exception problems can be matched directly through exception handling methods (specifically, the monitored insurance system can directly call exception handling methods), thereby reducing the repeated work performed to meet the same exception handling needs of different systems, realizing the management and reuse of exception handling methods by the monitored insurance system, and improving the reusability and configurability of exception handling; in this way, both can reduce the configuration of manpower and material resources.

[0024] Furthermore, the establishment of a connection with the monitored insurance system through the preset monitoring system includes: A distributed network connection is established with the network communication module in the preset monitoring system through the interface of the monitored insurance system, so as to implement monitoring services for the monitored insurance system through the preset monitoring system.

[0025] Understandably, in order to achieve distributed network connection between the monitoring system and multiple monitored insurance systems, it is necessary to build a centralized monitoring architecture based on JMX (Java Management Extensions) and RMI (Remote Method Invocation) protocols. The specific process includes network planning, security design, communication mechanism and data interaction logic. JMX+RMI combination: JMX provides standardized management interfaces (such as memory, thread, and GC monitoring). RMI implements remote calls of Java objects in a distributed environment and is used to transmit JMX requests. The monitoring center initiates the request. More specifically, the preset monitoring system calls the JMX interface of the monitored insurance system through the RMI client (such as getMemoryPoolMXBeans()); the request contains the target node ID, operation name (such as getHeapMemoryUsage) and authentication credentials; the implementation methods of the network communication module include 1. Protocol stack: TCP long connection (keep-alive mechanism 30s) + Protobuf serialization; 2. Flow control: The token bucket algorithm limits the maximum throughput of a single node (100,000 / second); 3. Encrypted transmission: The national secret SM4 algorithm encrypts key fields, and RSA exchanges keys; at the same time, optimization strategies are set in the network communication module, including 1. Data sharding: Sharding by device fingerprint hash to ensure routing consistency of homologous data; 2. Resume transmission: Record the data location index of the last 6 hours, and synchronize incrementally after reconnection; 3. Intelligent degradation: Automatically switch to the MQTT protocol to ensure reachability when the network jitters; Monitoring service can refer to the abnormal monitoring of the monitored insurance system through the preset monitoring system, and can implement corresponding processing methods.

[0026] Furthermore, before monitoring the insurance data object to be monitored using the insurance data monitoring model accumulated in the preset monitoring system, the method further includes: Automatically capture multiple operating indicators; Sampling insurance data corresponding to the operating indicators according to a preset sampling strategy; Marking the insurance data according to a preset marking strategy to obtain marked insurance data; After processing the marked insurance data in accordance with a preset accumulation strategy in terms of time dimension and coverage principle, accumulated insurance data is obtained; An initial insurance data monitoring model is trained and accumulated based on the accumulated insurance data to obtain at least one insurance data monitoring model.

[0027] Understandably, dynamic tracking is performed on the monitored insurance system. This allows JMX MBeans to automatically capture over 300 operational metrics, such as thread pool activity and SQL execution time. Preset sampling strategies can sample at intervals based on different time periods, such as 1-second intervals during peak periods and 60-second intervals during off-peak periods. Insurance data can be data collected from the monitored insurance system corresponding to operational metrics, such as thread pool activity and SQL execution time. Preset labeling strategies can automatically associate business scenarios with corresponding tags, such as those for "Double Eleven" and "New Institution Launch," where each tag corresponds to a category. Preset accumulation strategies can refer to the standards for accumulating insurance data (automatically extracting the latest data to update the model training set at a fixed time each day). These can be divided into time dimensions and coverage principles. The time dimension can refer to data containing at least two complete business cycles (such as the annual renewal peak). Scenario coverage can refer to covering more than 90% of the core business system's operational scenarios. The initial insurance data monitoring model can be a model structure that has not been trained on data, such as the initial Prophet. , GRU, and Isolation Forest, among which Prophet is used for time series prediction, can automatically model trends, seasonality and holiday effects, is robust to missing values ​​and outliers, and is suitable for data with obvious trends and periodicity. GRU is a recurrent neural network (RNN) variant used to process sequence data, which can effectively capture dependencies in long sequences, alleviate the gradient disappearance problem, and is suitable for processing time series data. IsolationForest It is used for anomaly detection. By building a random tree model, it "isolates" the data to identify unusual points. It has good robustness for high-dimensional data and large data sets, high computational efficiency, and is suitable for real-time anomaly detection. The insurance data monitoring model is the data model accumulated after training the above model, which can process data at different data layers. In this embodiment, the process builds a highly adaptable monitoring system, constructs monitoring modeling through a data closed loop, upgrades traditional passive response monitoring to active predictive risk control, and helps insurance institutions achieve cost reduction and efficiency improvement in the three dimensions of product innovation, operational efficiency, and compliance and safety.

[0028] Furthermore, the training and accumulation of the initial insurance data monitoring model based on the accumulated insurance data to obtain at least one insurance data monitoring model includes: Filter time series feature data through preset feature engineering; Extracting target feature data from the time series feature data through preset feature engineering; A collection model consisting of an insurance data monitoring model is trained based on the target feature data.

[0029] Understandably, the preset feature engineering can directly determine whether the model can capture the dynamic changes of risks (such as claims trends, user behavior attenuation), specifically including screening targets, among which the screening target can be to identify time-dependent features (such as "claim rate on the Nth day after the policy takes effect"), extract trend mutation signals (such as the car insurance claim rate continues to rise within 3 months after the policy adjustment), reduce data redundancy, remove high autocorrelation features (such as "the correlation between the "claim amount on the day" and "the claim amount yesterday" is greater than 0.9), compress low information density sequences (such as user login logs with no fluctuations for 30 consecutive days), convert original time series data (such as daily records) into statistical summaries (such as 7-day moving average, standard deviation), reduce computational overhead, and retain key time windows (such as "behavior 30 days before insurance" has a much greater impact on renewal rate than the historical average); extract target information from time series feature data. The target feature data can be further determined feature data, and the data can be filtered through the methodologies in the preset feature engineering and insurance scenario adaptation. The methodologies can include autocorrelation analysis, stationarity test, information entropy and sequence length truncation. The insurance scenario adaptation can include retaining time series features that are strongly correlated with fraud labels, discarding features that highly overlap with normal claim sequences, using survival analysis to screen time series features that have a significant impact on renewal time, combining Markov chain modeling to model user state transition probabilities, extracting "state residence time" features, retaining highly interpretable time series features (such as "the average annual growth rate of claims ratio in the past five years"), and removing features that overfit history (such as short-term features that are only valid during a specific policy period); the insurance data monitoring model can be the data set model mentioned above, and the training data set used in the model can be the target feature data obtained by screening; In this embodiment, through automated feature engineering, a rich set of time series features can be quickly constructed, significantly improving the model performance in scenarios such as insurance risk control and user segmentation.

[0030] Furthermore, before calling the processing method corresponding to the operating situation according to the exception handling library in the preset monitoring system to solve the operating situation, the method further includes: Accumulate corresponding insurance data for each operating indicator according to different severity levels; After determining the processing method and data waveform corresponding to the insurance data, the processing method and data waveform corresponding to the insurance data are accumulated in the abnormality processing library in the preset monitoring system.

[0031] Understandably, the severity level can be divided according to the level classification determined in the monitoring scenario, and each operating indicator can accumulate insurance data with multiple abnormal fluctuations; the data waveform can be divided into stable fluctuations, trend rise, seasonal pulse, sudden spike, continuous trough and periodic decay, etc. The processing method can be routine monitoring without intervention, triggering actuarial pricing review, increasing holiday resource allocation, immediately launching fraud investigation, triggering user retention strategy and adjusting renewal preferential policies, etc.; the exception handling library can be a pre-built module in the preset monitoring system, through which the corresponding processing method can be quickly matched from the knowledge graph in the later stage; In this embodiment, by presetting the processing methods accumulated in the exception processing library of the insurance monitoring system, closed-loop management from dynamic waveform recognition to accumulation of processing solutions can be achieved, which significantly improves the risk response efficiency.

[0032] Furthermore, the calling of a processing method corresponding to the operating condition according to the exception handling library in the preset monitoring system to resolve the operating condition includes: Calculating the similarity between the waveform of the insurance data object to be monitored in the operating condition and the waveform of the data in the exception processing library; The relationship between the set threshold and the similarity is used to trigger a processing method corresponding to the data waveform.

[0033] Understandably, the similarity between the waveform and the data waveform in the exception processing library can be measured by calculating the minimum regularized path distance of the two time series through DTW. This can include phase differences: such as whether the peak of claims settlement occurs earlier or later, and length differences: such as historical data is monthly statistics and current data is weekly statistics. The threshold can be a pre-set alarm threshold. When the similarity reaches this threshold, it means that the corresponding processing method needs to be determined to solve the current exception problem. In this embodiment, waveforms that are highly similar to historical anomalies can be effectively identified, assisting the risk control team in quickly responding to potential risk issues.

[0034] Furthermore, before calculating the similarity between the waveform of the insurance data object to be monitored and the data waveform in the exception processing library and triggering a processing method corresponding to the data waveform based on the relationship between a set threshold and the similarity, the method further includes: Setting the rule processing method in the exception handling library through a preset rule engine; The processing methods corresponding to the data waveforms in the exception processing library are stored through a preset knowledge graph.

[0035] Understandably, the pre-set rule engine is used for complex event processing and automated decision-making. This involves setting the rule processing methods in the exception handling library and the conditions under which each processing method is triggered, such as grouping by service type (claims, underwriting, customer service) and severity level (P0-P3). For example, if a service fails three consecutive heartbeats, the rule ServiceDownRule is triggered. Each node in the knowledge graph can store processing methods, so pre-stored processing methods can be queried. For example, if a query shows that the database auth_db that the service depends on is undergoing maintenance, the pre-set rule engine can execute a link switching action to route traffic to the backup database (that is, the pre-set rule engine can trigger the corresponding processing method). In this embodiment, full-link automation from rule matching to root cause analysis can be achieved, and the processing method can be quickly determined from the exception processing library.

[0036] It should be noted that the preset monitoring system also includes a data sampling information library, which is built using a distributed time series database (such as TDengine), deploying a 3-node cluster to ensure high availability, and designing a two-level storage strategy: hot data is retained for 7 days (SSD storage), and cold data is archived to object storage (retention for 5 years). The data model is defined: the collection frequency (100ms~1min), data accuracy (float64), and label system (three-level labels for environment / application / host) are configured for each indicator. The processing mechanism may include streaming compression: using the Gorilla compression algorithm to reduce storage space by 80%; sampling calibration: eliminating burr data through the sliding window algorithm (window size 5min); data completion: using cubic spline interpolation to fill missing values ​​at abnormal breakpoints.

[0037] The preset monitoring system also includes a data collection module, which uses a collector deployment: Filebeat + Telegraf is deployed in the Kubernetes DaemonSet mode. Resource isolation: independent CPU cgroups are divided to limit the maximum resource usage (CPU 15% / memory 2GB). Adaptive sampling: the collection frequency is dynamically adjusted according to the system load (1s~60s). The processing flow may include format standardization: using Grok expressions to parse 300+ log formats; data cleaning: filtering test environment data and heartbeat messages; priority classification: business indicators > system indicators > debug logs.

[0038] The preset monitoring system also includes a page display module, which uses a visualization engine: a configurable large screen built based on ECharts, intelligent analysis: integrated natural language query (supports queries such as "TOP5 abnormal services in the past 2 hours"), and permission control: granular to the field level (for example, only Shanghai data center operations and maintenance can see local indicators). Key functions include root cause topology diagrams: automatic generation of service dependency anomaly propagation paths, prediction views: LSTM models predict indicator trends for the next 1 hour, and correlation analysis: displaying configuration changes and release records at the time of the anomaly.

[0039] The preset monitoring system also includes a monitoring database, which uses tiered storage. The hot layer: Redis cluster caches real-time status data (TTL 2 hours), the warm layer: TiDB stores aggregated indicators (1-minute granularity), and the cold layer: HDFS archives raw data. Index optimization: establishes a combined index for high-frequency query fields (such as time + service name). The operation and maintenance mechanism includes automatic table partitioning: partitioning by double timestamps (collection time + storage time); data governance: regularly performs data quality checks (null value rate / outlier detection); capacity warning: automatically triggers the expansion approval process when the storage capacity reaches 80%; typical workflow example (abnormal auto insurance quotation): the data collection module captures sudden increases in quotation delays; the monitoring model library identifies deviations from the baseline (exceeding the dynamic threshold of 173%); the exception handling library matches the knowledge graph and recommends "checking the Redis cluster"; the page display module highlights the abnormal Redis node in the Shanghai data center; the configuration library automatically issues a restart command and verifies the recovery.

[0040] In summary, the above provides an insurance information monitoring method, which establishes a connection with the monitored insurance system through a preset monitoring system; obtains the insurance data object to be monitored from the monitored insurance system; monitors and processes the insurance data object to be monitored through the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; judges the operation status of the monitored insurance system based on the monitoring operation data, and calls the processing method corresponding to the operation status according to the exception handling library in the preset monitoring system to solve the operation status. In this solution, 1. By building a monitoring model library within a pre-set monitoring system, the system can directly create monitoring objects using insurance data objects and model objects (specifically, the monitored insurance system can directly call the insurance data objects and model objects), reducing the duplication of work required to create monitoring objects by modifying configuration files and enabling management and reuse of monitoring data and models. By building an exception handling library within a pre-set monitoring system, the system can directly match exception problems using exception handling methods (specifically, the monitored insurance system can directly call the exception handling methods), reducing the duplication of work required to handle the same exception handling needs of different systems, enabling management and reuse of exception handling methods within the monitored insurance system and improving the reusability and configurability of exception handling. Thus, both reduce the allocation of manpower and material resources. 2. The monitored insurance data objects to be monitored are obtained from the monitored insurance system, enabling the monitored insurance system to independently select monitoring data. The monitored insurance system can then define monitoring data objects and monitoring details based on its own operational needs. By distinguishing between critical and non-critical data during operation, it can focus on critical data, thereby improving monitoring efficiency. 3. Call the processing method corresponding to the operating situation according to the exception handling library in the preset monitoring system to solve the operating situation, and provide a corresponding relationship between the custom exception threshold (by setting the similarity threshold) and the exception handling method, so that the monitored insurance system can handle the actual exception situation conveniently, and automatically call the exception handling method when an exception occurs by setting or modifying the threshold control, thereby reducing manual participation in the exception handling process, transforming the original passive processing into active processing, improving the flexibility of exception handling, accelerating the speed of locating exceptions, shortening the fault recovery time, and ensuring the stability of system operation.

[0041] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0042] In one embodiment, an insurance information monitoring device is provided, which corresponds to the insurance information monitoring method in the above embodiment. Figure 3As shown, the insurance information monitoring device includes a creation module 11, an acquisition module 12, a monitoring and processing module 13, and a judgment module 14. The functional modules are described in detail as follows: Establishing module 11, used to establish a connection with the monitored insurance system through a preset monitoring system; An acquisition module 12, configured to acquire the insurance data object to be monitored from the monitored insurance system; A monitoring processing module 13 is configured to perform monitoring processing on the insurance data object to be monitored using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; The judgment module 14 is used to judge the operation status of the monitored insurance system according to the monitoring operation data, and call a processing method corresponding to the operation status according to the exception processing library in the preset monitoring system to solve the operation status.

[0043] Furthermore, the establishment module includes: The establishment submodule is used to establish a distributed network connection with the network communication module in the preset monitoring system through the interface of the monitored insurance system, so as to realize the monitoring service of the monitored insurance system through the preset monitoring system.

[0044] Furthermore, the insurance information monitoring device further includes: Capture module, used to automatically capture various operating indicators; A sampling module, configured to sample insurance data corresponding to the operating indicators according to a preset sampling strategy; a marking module, configured to mark the insurance data according to a preset marking strategy to obtain marked insurance data; a processing module, configured to process the marked insurance data in accordance with a preset accumulation strategy in terms of time dimension and coverage principle to obtain accumulated insurance data; The first accumulation module is configured to train and accumulate an initial insurance data monitoring model based on the accumulated insurance data to obtain at least one insurance data monitoring model.

[0045] Furthermore, the first accumulation module includes: The screening submodule is used to screen time series feature data through preset feature engineering; An extraction submodule, configured to extract target feature data from the time series feature data through a preset feature engineering; The training submodule is used to train a set model consisting of an insurance data monitoring model based on the target feature data.

[0046] Furthermore, the insurance information monitoring device further includes: The second accumulation module is used to accumulate corresponding insurance data for each operating indicator according to different severity levels; The third accumulation module is used to determine the processing method and data waveform corresponding to the insurance data, and then accumulate the processing method and data waveform corresponding to the insurance data into the abnormality processing library in the preset monitoring system.

[0047] Furthermore, the judgment module includes: a calculation submodule, configured to calculate the similarity between the waveform of the insurance data object to be monitored in the operation condition and the waveform of the data in the exception processing library; The trigger submodule is used to trigger a processing method corresponding to the data waveform according to the relationship between the set threshold and the similarity.

[0048] Furthermore, the insurance information monitoring device further includes: A setting module, used to set the rule processing method in the exception processing library through a preset rule engine; A storage module is used to store the processing methods corresponding to the data waveforms in the exception processing library through a preset knowledge graph.

[0049] The specific definition of the insurance information monitoring device can be found in the definition of the insurance information monitoring method above and will not be further elaborated here. Each module in the aforementioned insurance information monitoring device may be implemented in whole or in part via software, hardware, or a combination thereof. Each module may be embedded in or independent of a processor within a computer device in hardware form, or may be stored in a computer device memory in software form, allowing the processor to call and execute the corresponding operations of each module.

[0050] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, memory, network interface and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. The database of the computer device is used to store data involved in the insurance information monitoring method. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, an insurance information monitoring method is implemented.

[0051] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the insurance information monitoring method in the above embodiment are implemented, such as Figure 2 Steps S10 to S20 shown; Alternatively, when the processor executes the computer program, the functions of the modules / units of the insurance information monitoring device in the above embodiment are realized, such as Figure 3 The functions of modules 11 and 12 are shown in FIG.

[0052] In one embodiment, a computer-readable storage medium is provided on which a computer program is stored. When the computer program is executed by a processor, the steps of the insurance information monitoring method in the above embodiment are implemented, such as Figure 2 Steps S10 to S20 shown; Alternatively, when the computer program is executed by the processor, the functions of the modules / units of the insurance information monitoring device in the above embodiment are realized, such as Figure 3 The functions of modules 11 and 12 are shown in FIG.

[0053] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware using a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0054] Those skilled in the art will clearly understand that for the sake of convenience and brevity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0055] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. A method for monitoring insurance information, characterized in that: include: Establishing a connection with the monitored insurance system through a preset monitoring system; Acquiring an insurance data object to be monitored from the monitored insurance system; Performing monitoring processing on the insurance data object to be monitored by using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; The operation status of the monitored insurance system is determined according to the monitoring operation data, and a processing method corresponding to the operation status is called according to the exception processing library in the preset monitoring system to solve the operation status.

2. The insurance information monitoring method according to claim 1, characterized in that: The establishment of a connection with the monitored insurance system through the preset monitoring system includes: A distributed network connection is established with the network communication module in the preset monitoring system through the interface of the monitored insurance system, so as to implement monitoring services for the monitored insurance system through the preset monitoring system.

3. The insurance information monitoring method according to claim 1, characterized in that: Before monitoring the insurance data object to be monitored using the insurance data monitoring model accumulated in the preset monitoring system, the method further includes: Automatically capture multiple operating indicators; Sampling insurance data corresponding to the operating indicators according to a preset sampling strategy; Marking the insurance data according to a preset marking strategy to obtain marked insurance data; After processing the marked insurance data in accordance with a preset accumulation strategy in terms of time dimension and coverage principle, accumulated insurance data is obtained; An initial insurance data monitoring model is trained and accumulated based on the accumulated insurance data to obtain at least one insurance data monitoring model.

4. The insurance information monitoring method according to claim 3, characterized in that: The step of training and accumulating the initial insurance data monitoring model based on the accumulated insurance data to obtain at least one insurance data monitoring model includes: Filter time series feature data through preset feature engineering; Extracting target feature data from the time series feature data through preset feature engineering; A collection model consisting of an insurance data monitoring model is trained based on the target feature data.

5. The insurance information monitoring method according to claim 1, characterized in that: Before calling the processing method corresponding to the operating situation according to the exception processing library in the preset monitoring system to solve the operating situation, the method further includes: Accumulate corresponding insurance data for each operating indicator according to different severity levels; After determining the processing method and data waveform corresponding to the insurance data, the processing method and data waveform corresponding to the insurance data are accumulated in the abnormality processing library in the preset monitoring system.

6. The insurance information monitoring method according to claim 1, characterized in that: The calling of a processing method corresponding to the operating condition according to the exception handling library in the preset monitoring system to resolve the operating condition includes: Calculating the similarity between the waveform of the insurance data object to be monitored in the operating condition and the waveform of the data in the exception processing library; The relationship between the set threshold and the similarity is used to trigger a processing method corresponding to the data waveform.

7. The insurance information monitoring method according to claim 6, characterized in that: Before calculating the similarity between the waveform of the insurance data object to be monitored and the data waveform in the exception processing library, and triggering a processing method corresponding to the data waveform based on the relationship between a set threshold and the similarity, the method further includes: Setting the rule processing method in the exception handling library through a preset rule engine; The processing methods corresponding to the data waveforms in the exception processing library are stored through a preset knowledge graph.

8. An insurance information monitoring device, characterized in that: include: An establishment module for establishing a connection with a monitored insurance system through a preset monitoring system; An acquisition module, configured to acquire the insurance data object to be monitored from the monitored insurance system; A monitoring processing module, configured to perform monitoring processing on the insurance data object to be monitored using the insurance data monitoring model accumulated in the preset monitoring system to obtain monitoring operation data; The judgment module is used to judge the operation status of the monitored insurance system according to the monitoring operation data, and call the processing method corresponding to the operation status according to the exception processing library in the preset monitoring system to solve the operation status.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the insurance information monitoring method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the insurance information monitoring method according to any one of claims 1 to 7 is implemented.

Images