A dynamic diversion method and system for insurance underwriting process versions based on user segmentation

Abstract

The invention relates to the technical field of insurance businesses, in particular to an insurance flow version dynamic shunting method and system based on user grouping, and overcomes the defects of long experimental period and response delay of a traditional static A / B test by collecting conversion rate data in real time and dynamically adjusting the user shunting proportion. The high-conversion-rate version flow can be amplified in time, and the experiment efficiency and the insurance success rate are remarkably improved; meanwhile, flow distribution is performed by combining user grouping and a dynamic distribution proportion, so that personalized recommendation based on user attributes and real-time intentions is realized, dynamic adaptability is enhanced, and a technical basis is provided for deep fusion of insurance service rules through a grouping mechanism, so that the efficiency is improved, and the user experience is improved. And the accuracy and compliance of business operation are ensured.

Description

(19) State Intellectual Property Office (12) Invention Patent Application (10) Application Publication Number (43) Application Publication Date (21) Application Number 202511744721.8 (22) Application Date 2025.11.25 (71) Applicant Beijing Qingsong Yikang Information Technology Co., Ltd. Address Room 716, 7th Floor, Building 2, No. 28, Andingmen East Street, Dongcheng District, Beijing 100007 (72) Inventors Yang Yin Wu Wanwei Liu Junshi Li Zhiguo Zhao Jie Zhou Zhimin Gao Xin Jia Haiyong Ruan Sheng Zhu Xuexin Zhao Na Chen Guangyao Liu Yuanchuang (74) Patent Agency Beijing Huaxia Taihe Intellectual Property Agency Co., Ltd. 11662 Patent Attorney Yang Jun (51) Int.Cl. G06Q 40 / 08 (2012.01) G06Q 30 / 0201 (2023.01) G06F 16 / 9535 (2019.01) G06F 16 / 9537(2019.01) (54) Invention Title: A Method and System for Dynamic Diversion of Insurance Application Process Versions Based on User Segmentation (57) Abstract: This application relates to the field of insurance business technology, specifically to a method and system for dynamic diversion of insurance application process versions based on user segmentation. This application overcomes the shortcomings of traditional static A / B testing, such as long experimental cycles and response delays, by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio. It can amplify the traffic of high conversion rate versions in a timely manner, significantly improving experimental efficiency and insurance application success rate. At the same time, by combining user segmentation with dynamic diversion ratio for process allocation, it not only realizes personalized recommendations based on user attributes and real-time intent, enhancing dynamic adaptability, but also provides a technical basis for deep integration of insurance business rules through the segmentation mechanism, thereby improving efficiency while ensuring the accuracy and compliance of business operations. Claims 3 pages, Description 16 pages, Drawings 2 pages, CN 121504635 A 2026.02.10 CN 1 21 50 46 35 A 1. A method for dynamic diversion of insurance application process versions based on user segmentation, characterized in that it includes: responding to a user's insurance application request, acquiring user attribute data and real-time behavior data; classifying the user into a predefined user segment according to the user attribute data and the real-time behavior data; assigning a corresponding insurance application process version to the user based on the user segment and the currently effective user diversion ratio; collecting and statistically analyzing the conversion rate data of each insurance application process version in real time, and determining whether it meets the preset dynamic adjustment conditions; if it is determined that the dynamic adjustment conditions are met, calculating and updating the user diversion ratio according to the conversion rate data, and assigning an insurance application process version to subsequent users according to the updated user diversion ratio. 2. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, based onBased on the user attribute data and the real-time behavior data, the user is divided into predefined user groups, including: obtaining the geographical location attribute from the user attribute data and matching the geographical location attribute with a preset high-risk area list, wherein the high-risk area list is pre-configured based on the strictness of insurance regulatory policies in each region; if the match is successful, the user is determined to be in a preset high-risk area, and the first grouping logic is executed: obtaining the form filling progress from the real-time behavior data, wherein the form filling progress is obtained by tracking the ratio of the number of required form fields completed by the user in the insurance application process version to the total number of required form fields; determining whether the form filling progress is greater than a first preset threshold; if the form filling progress is greater than the first preset threshold, the user is divided into a user group labeled as high-risk and high-intention; if the form filling progress is less than or equal to the first preset threshold, the user is divided into a user group labeled as high-risk. If a match fails, it is determined that the user is not in a high-risk area, and the second grouping logic is executed: obtain the page dwell time and form filling progress from the real-time behavior data; determine whether the page dwell time is greater than a second preset threshold, and simultaneously determine whether the form filling progress is greater than a third preset threshold; if it is determined that the page dwell time is greater than the second preset threshold and the form filling progress is greater than the third preset threshold, then the user is divided into the user group labeled as high intent; if it is determined that the page dwell time is less than or equal to the second preset threshold and / or the form filling progress is less than or equal to the third preset threshold, then the user is divided into the user group labeled as ordinary intent. 3. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that determining whether a preset dynamic adjustment condition is met includes: obtaining the current system time and retrieving a preset adjustment period, wherein the adjustment period defines the time interval for dynamic adjustment; calculating the time difference between the current time and the time point of the last successful dynamic adjustment; determining whether the time difference reaches or exceeds the adjustment period; if the time difference reaches or exceeds the adjustment period, it is determined that the dynamic adjustment condition is met. 4. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, calculating and updating the user diversion ratio based on the conversion rate data includes: obtaining the real-time conversion rate of each insurance application process version within the current statistical window, wherein the real-time conversion rate is the ratio of the number of users who successfully completed the insurance application process version to the total number of users accessing the insurance application process version; determining the insurance type weight based on the type of insurance product carried by each insurance application process version; and determining the weighted conversion rate using the weighted conversion rate calculation formula based on the real-time conversion rate and the insurance type weight of each insurance application process version.5. The dynamic diversion method for insurance application process versions based on user segmentation according to claim 4, characterized in that, the insurance product weight is determined according to the insurance product type carried by each insurance application process version, including: obtaining the basic weight corresponding to the insurance product type according to the insurance product type, based on a predefined insurance product weight configuration table; obtaining the current system time and determining whether the current time is within a preset target period, wherein the target period includes holidays, large-scale promotional activities, and high-value time periods of the day; if the current time is determined to be within the preset target period, then the final effective insurance product weight is determined based on the current time and a preset time period-weight mapping table, wherein the time period-weight mapping table stores the mapping relationship between a specific time period and the corresponding insurance product weight in the form of key-value pairs; if the current time is determined not to be within the preset target period, then the basic weight is used as the final effective insurance product weight. 6. The method for dynamic traffic allocation of insurance application process versions based on user segmentation according to claim 1, characterized in that, based on the user segmentation and the currently effective user traffic allocation ratio, assigning a corresponding insurance application process version to the user includes: determining one or more candidate insurance application process versions according to the segmentation tag corresponding to the user segmentation; obtaining the currently effective user traffic allocation ratio, wherein the user traffic allocation ratio defines the proportion of user traffic that each candidate version should be allocated at the current moment; and using a consistent hashing algorithm to bucket user identifiers according to the currently effective user traffic allocation ratio, thereby assigning a unique target insurance application process version to the current user from the candidate insurance application process versions. 7. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, before allocating insurance application process versions to subsequent users according to the updated user diversion ratio, the method further includes: after calculating a preliminary user diversion ratio adjustment plan based on the conversion rate data, identifying one or more target insurance application process versions with a planned increase in diversion ratio from the plan; for each target insurance application process version with a planned increase, executing a real-time compliance verification process, specifically including: obtaining the user ID of the current user to be diverted, and querying the corresponding user attribute data based on the user ID, wherein the user attribute data includes at least geographical location and age; using the user's geographical location, age, and the identifier of the target insurance application process version as input, calling the compliance verification service, and verifying by querying a pre-set compliance rule library to determine whether the target insurance application process version meets all mandatory compliance requirements for the corresponding user's location, insurance type, and user attributes;If the verification passes, the target insurance application process version is determined to meet the compliance requirements, and the process of increasing the diversion ratio for it is allowed; if the verification fails, the target insurance application process version is determined to not meet the compliance requirements, and the process of increasing the diversion ratio for the target insurance application process version is cancelled. 8. A dynamic diversion system for insurance application process versions based on user segmentation, characterized in that it comprises: (Claims 2 / 3, page 3, CN 121504635 A) A data acquisition module, used to acquire user attribute data and real-time behavior data in response to a user's insurance application request; a user segmentation module, used to divide the user into predefined user segments according to the user attribute data and the real-time behavior data; a version allocation module, used to allocate a corresponding insurance application process version to the user based on the user segment and the currently effective user diversion ratio; a data collection and statistics module, used to collect and statistically analyze the conversion rate data of each insurance application process version in real time; a strategy decision module, used to determine whether preset dynamic adjustment conditions are met; when it is determined that the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated according to the conversion rate data; wherein, the version allocation module is further used to allocate insurance application process versions to subsequent users according to the updated user diversion ratio. 9. An electronic device, comprising a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other via the communication bus, characterized in that: the memory is used to store a computer program; the processor is used to execute the dynamic diversion method for the insurance application process version based on user segmentation as described in any one of claims 1 to 7 by running the computer program stored in the memory. 10. A computer-readable storage medium, characterized in that: the storage medium stores a computer program, wherein the computer program is configured to execute the dynamic diversion method for the insurance application process version based on user segmentation as described in any one of claims 1 to 7 when running. Claims 3 / 3 Page 4 CN 121504635 A Dynamic Diversion Method and System for Insurance Application Process Version Based on User Segmentation Technical Field

[0001] This application relates to the technical field of insurance business, specifically to a dynamic diversion method and system for the insurance application process version based on user segmentation. Background Technology

[0002] In the field of insurance technology, improving the conversion rate of the insurance application process is one of the core technical goals. The key lies in accurately identifying user needs and dynamically optimizing page redirection paths. Current mainstream technical solutions mainly rely on two types: one is a static A / B testing framework, which allocates user traffic to different process versions according to a preset fixed ratio, and then uniformly evaluates the effect after the experimental period ends; the other is an engine based on predefined IF-THEN rules, which recommends redirection paths based on user static attributes or historical behavior data.

[0003] However, these existing technical solutions have obvious drawbacks. First, the experimental efficiency is low: static A / B testing requires waiting for a complete experimental cycle (usually several days or even weeks) to draw conclusions, and cannot respond in real time to the process version that shows a high conversion rate during the experiment, resulting in missed best promotion opportunities and potential loss of sales. Second, the dynamic adaptability is poor: for the long decision chain scenario unique to insurance products, the static rule-based solution is difficult to capture real-time changes in user behavior intent, resulting in a serious mismatch between the recommended jump path and the user's dynamic needs. Finally, the business targeting is weak: general technical solutions fail to deeply integrate the special characteristics of insurance business. For example, they do not consider the differentiated value of different insurance types, and they lack a verification mechanism for the strict regional compliance requirements of the insurance industry, resulting in insufficient business adaptability and compliance risks.

[0004] Therefore, the existing technology has three core bottlenecks: delayed experimental response, rigid dynamic strategies, and insufficient business adaptability. There is an urgent need for a method for diverting insurance application process versions that can deeply integrate insurance business rules and support real-time feedback and dynamic adjustment, so as to achieve the precision, compliance and efficiency of the optimization process.

[0005] The purpose of this application is to provide a method and system for dynamic diversion of insurance application process versions based on user segmentation, so as to solve the problems of low experimental efficiency and response delay in the optimization of insurance application process versions by static A / B testing in the prior art.

[0006] In order to achieve the above objective, the technical solution adopted by this application is as follows: According to one aspect of the embodiment of this application, a method for dynamic diversion of insurance application process versions based on user segmentation is provided, including: in response to a user's insurance application request, obtaining user attribute data and real-time behavior data; dividing the user into a predefined user segment according to the user attribute data and real-time behavior data; assigning a corresponding insurance application process version to the user based on the user segment and the currently effective user diversion ratio; collecting and statistically analyzing the conversion rate data of each insurance application process version in real time, and determining whether a preset dynamic adjustment condition is met; if it is determined that the dynamic adjustment condition is met, calculating and updating the user diversion ratio according to the conversion rate data, and assigning an insurance application process version to subsequent users according to the updated user diversion ratio.

[0007] Based on the above technical means, by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio, the inherent defects of traditional static A / B testing, such as long experimental cycles and response delays, are effectively overcome. This allows for timely amplification of traffic for high-conversion-rate versions, thereby significantly improving experimental efficiency and overall conversion rate. Simultaneously, by combining user segmentation mechanisms with dynamic diversion strategies, personalized process recommendations based on user attributes and real-time behavioral intentions are achieved, enhancing the system's adaptability to dynamic user needs. This also provides a basis for deep integration with insurance business rules (such as regional compliance verification and differentiated weighting of insurance types).This provides the technical foundation, thereby ensuring the accuracy and compliance of business processes while pursuing maximum conversion efficiency.

[0008] Further, based on user attribute data and real-time behavior data, users are divided into predefined user groups, including: obtaining the geographical location attribute in the user attribute data and matching the geographical location attribute with a preset high-risk area list, wherein the high-risk area list is pre-configured according to the strictness of insurance regulatory policies in each region; if the match is successful, it is determined that the user is in a preset high-risk area, and the first grouping logic is executed: obtaining the form filling progress in the real-time behavior data, wherein the form filling progress is obtained by tracking the ratio of the number of required form fields completed by the user in the insurance application process version to the total number of required form fields; determining whether the form filling progress is greater than a first preset threshold; if it is determined that the form filling progress is greater than the first preset threshold, the user is divided into groups tagged as high-risk. The system first segments users by intention. If the form completion progress is less than or equal to a first preset threshold, the user is assigned to the high-risk user group. If no match is found, the user is determined not to be in a high-risk area, and the second segmentation logic is executed: The system retrieves page dwell time and form completion progress from real-time behavioral data; it determines whether the page dwell time is greater than a second preset threshold and whether the form completion progress is greater than a third preset threshold. If the page dwell time is greater than the second preset threshold and the form completion progress is greater than the third preset threshold, the user is assigned to the high-intent user group. If the page dwell time is less than or equal to the second preset threshold and / or the form completion progress is less than or equal to the third preset threshold, the user is assigned to the ordinary-intent user group.

[0009] Based on the above technical means, by deeply integrating static user attributes such as geographical location with dynamic behavioral intentions such as form filling progress and page dwell time, a refined real-time grouping model is constructed, which can automatically identify and distinguish user groups with different characteristics such as "high compliance requirements" and "high conversion potential". On the one hand, this mechanism can ensure that users in high-risk areas are accurately guided to the insurance process version that includes mandatory compliance steps, thus avoiding regulatory risks from the source. On the other hand, it can prioritize the allocation of high-intent users to the fast channel with a better experience, effectively improving the conversion rate. Thus, while ensuring business compliance, it achieves simultaneous optimization of traffic allocation efficiency and accuracy.

[0010] Further, determining whether the preset dynamic adjustment conditions are met includes: obtaining the current system time and retrieving the preset adjustment period, wherein the adjustment period defines the time interval for dynamic adjustment; calculating the time difference between the current time and the time point of the last successful dynamic adjustment; determining whether the time difference reaches or exceeds the adjustment period; if the time difference reaches or exceeds the adjustment period, it is determined that the dynamic adjustment conditions are met.

[0011] According to the above technical means, by introducing an automated triggering mechanism based on a fixed time period, the dynamic adjustmentThe shift from an event-driven model reliant on manual intervention or data fluctuations to a systematic time-driven model enables the periodic and automated self-optimization of traffic diversion strategies. This not only completely avoids the response delay problem of traditional A / B testing, which requires waiting for the entire experimental cycle to conclude before drawing conclusions, but also allows the system to respond quickly to changes in conversion rates at a preset pace and promptly amplify high-potential version traffic, thereby significantly improving experimental efficiency and resource utilization. At the same time, it ensures the controllability, predictability, and stability of the system's adjustment behavior, avoiding frequent strategy oscillations caused by instantaneous fluctuations in real-time data.

[0012] Further, the user diversion ratio is calculated and updated based on the conversion rate data, including: obtaining the real-time conversion rate of each insurance application process version within the current statistical window, wherein the real-time conversion rate is the ratio of the number of users who successfully completed the insurance application process version to the total number of users accessing the insurance application process version; determining the insurance type weight based on the type of insurance product carried by each insurance application process version; determining the weighted conversion rate corresponding to each insurance application process version based on the real-time conversion rate and insurance type weight of each insurance application process version through the weighted conversion rate calculation formula; and determining the user diversion ratio corresponding to each insurance application process version through the diversion ratio formula based on the weighted conversion rate and the sum of the weighted conversion rates of all insurance application process versions.

[0013] Based on the above technical means, by introducing the insurance type weight coefficient and weighting its real-time conversion rate, a deeply integrated business value-oriented diversion ratio decision model is constructed, which enables the diversion strategy to not only respond to the conversion efficiency of each process version, but also accurately reflect the differentiated business value of different insurance products. This mechanism automatically maps the weighted conversion rate to the diversion ratio through a normalization algorithm, realizing the optimal allocation of traffic resources to the "high conversion + high value" combination, thereby ensuring the mathematical rationality and business adaptability of the diversion decision.

[0014] Further, the insurance product weight is determined according to the insurance product type carried by each insurance application process version, including: obtaining the basic weight corresponding to the insurance product type according to the insurance product type and the predefined insurance product weight configuration table; obtaining the current system time and determining whether the current time is within the preset target period, wherein the target period includes holidays, large-scale promotional activities, and high-value periods of the day; if it is determined that the current time is within the preset target period, then the final effective insurance product weight is determined based on the current time and the preset time period-weight mapping table, wherein the time period-weight mapping table stores the mapping relationship between a specific time period and the corresponding insurance product weight in the form of key-value pairs; if it is determined that the current time is not within the preset target period, then the basic weight is used as the final effective insurance product weight.

[0015] According to the above technical means, by introducing a dynamic coupling mechanism between the time dimension and business strategy, the static insuranceThe basic weights are intelligently associated with preset high-value time periods, promotional activity periods, and other time factors, realizing adaptive optimization of the diversion strategy in different time contexts; this mechanism can automatically identify key business periods and increase the weight priority of high-value insurance products, ensuring that traffic resources are tilted towards high-yield products during peak periods, thereby significantly improving the overall input-output ratio and resource utilization of marketing activities without manual intervention in the system, while enhancing the system's dynamic response capability to changes in the market environment and the accuracy of strategy execution.

[0016] Further, based on user groups and the currently effective user diversion ratio, the corresponding insurance process version is assigned to the user, including: determining one or more candidate insurance process versions according to the group tags corresponding to the user groups; obtaining the currently effective user diversion ratio, wherein the user diversion ratio defines the proportion of user traffic that each candidate version should be allocated at the current moment; according to the currently effective user diversion ratio, the user identifier is bucketed using a consistent hash algorithm, thereby assigning a unique target insurance process version to the current user from the candidate insurance process versions.

[0017] Based on the above technical means, by combining business-oriented user grouping rules with data-driven traffic distribution ratio strategies, while ensuring that users with high compliance requirements are accurately guided to the safe process, the uniformity, stability and consistency of traffic distribution and user sessions are achieved by using the consistent hashing algorithm. This mechanism not only ensures that the traffic distribution strategy strictly follows business compliance constraints and avoids regulatory risks caused by version misallocation, but also ensures the accuracy of experimental data and system load balancing of A / B testing through predictable and low-conflict hash bucketing technology, thereby maintaining high performance and high reliability of traffic distribution decisions under complex business rules.

[0018] Further, before allocating insurance process versions to subsequent users based on the updated user diversion ratio, the method further includes: after calculating a preliminary user diversion ratio adjustment plan based on conversion rate data, identifying one or more target insurance process versions with a planned increase in diversion ratio from the plan; for each target insurance process version with a planned increase, executing a real-time compliance verification process, specifically including: obtaining the user ID of the user to be diverted, and querying the corresponding user attribute data based on the user ID, wherein the user attribute data includes at least geographical location and age; using the user's geographical location, age, and the identifier of the target insurance process version as input, calling the compliance verification service, and verifying by querying the pre-set compliance rule library to determine whether the target insurance process version meets all mandatory compliance requirements for the corresponding user's location, insurance type, and user attributes; if the verification passes, it is determined that the target insurance process version meets the compliance requirements, and the diversion ratio increase operation is allowed; if the verification fails, it is determined that the target insurance process version does not meet the compliance requirements, and the diversion ratio increase for the target insurance process version is canceled.

[0019] Based on the aforementioned technical means, a rule-driven security protection mechanism was constructed by embedding a real-time compliance verification step before the execution of the traffic diversion strategy. This mechanism can automatically identify and block traffic diversion operations that violate regional regulatory policies or specific user restrictions based on user location, age, and target process characteristics. This mechanism upgrades compliance verification from offline checks to real-time blocking in advance, effectively avoiding the risk of regulatory penalties caused by misallocation of process versions. It also ensures agile optimization of the traffic diversion strategy within the business compliance framework through automated system judgment, achieving a balance between business security and experimental efficiency.

[0020] According to another aspect of the embodiments of this application, a dynamic diversion system for insurance application process versions based on user segmentation is also provided, including: a data acquisition module, used to acquire user attribute data and real-time behavior data in response to a user's insurance application request; a user segmentation module, used to divide users into predefined user segments according to user attribute data and real-time behavior data; a version allocation module, used to allocate a corresponding insurance application process version to a user based on the user segment and the currently effective user diversion ratio; a data collection and statistics module, used to collect and statistically analyze the conversion rate data of each insurance application process version in real time; a strategy decision module, used to determine whether preset dynamic adjustment conditions are met; when it is determined that the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated according to the conversion rate data; wherein, the version allocation module is also used to allocate insurance application process versions to subsequent users according to the updated user diversion ratio.

[0021] According to another aspect of the embodiments of this application, an electronic device is also provided, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus; wherein the memory is used to store a computer program; and the processor is used to execute the steps of the dynamic diversion method for the insurance application process version based on user segmentation in any of the above embodiments by running the computer program stored in the memory.

[0022] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein the storage medium stores a computer program, wherein the computer program is configured to execute the steps of the dynamic diversion method for the insurance application process version based on user segmentation in any of the above embodiments when running.

[0023] The beneficial effects of this application: This application overcomes the shortcomings of traditional static A / B testing, such as long experimental cycles and response delays, by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio. It can promptly amplify the traffic of high-conversion versions, significantly improving experimental efficiency and insurance success rate. At the same time, by combining user grouping with dynamic diversion ratio for process allocation, it not only realizes personalized recommendations based on user attributes and real-time intent, enhancing dynamic adaptability, but also facilitates deep integration with the insurance industry through the grouping mechanism.The service rules provide a technical foundation, thereby improving efficiency while ensuring the accuracy and compliance of business operations.

[0024] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0025] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 is a schematic diagram of the hardware environment of an optional user-segmented insurance process version dynamic diversion method provided in an embodiment of this application; Figure 2 is a flowchart of an optional user-segmented insurance process version dynamic diversion method provided in an embodiment of this application; Figure 3 is a structural block diagram of an optional user-segmented insurance process version dynamic diversion system provided in an embodiment of this application; Figure 4 is a structural block diagram of an optional electronic device provided in an embodiment of this application. Specification 4 / 16 pages 8 CN 121504635 A Detailed Description

[0027] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that the terms "first", "second", etc. in the specification, claims and the above-mentioned drawings of the present application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0029] According to one aspect of an embodiment of this application, a dynamic diversion method for insurance application process versions based on user segmentation is provided. Optionally, in this embodiment, the above-described dynamic diversion method for insurance application process versions based on user segmentation can be applied to a hardware environment consisting of terminals and servers. The server connects to the terminal via a network and can be used to provide services to the terminal.The client installed on the terminal provides services and can set up a database on the server or independently of the server to provide data storage services for the server.

[0030] The above-mentioned network may include, but is not limited to, at least one of the following: wired network, wireless network. The above-mentioned wired network may include, but is not limited to, at least one of the following: wide area network, metropolitan area network, local area network. The above-mentioned wireless network may include, but is not limited to, at least one of the following: WIFI (Wireless Fidelity), Bluetooth. The terminal may not be limited to PC, mobile phone, tablet computer, etc.

[0031] The dynamic diversion method of the insurance application process based on user grouping in this application embodiment can be executed by the server, or by the terminal, or by both the server and the terminal. Among them, the terminal can also execute the dynamic diversion method of the insurance application process based on user grouping in this application embodiment by the client installed on it.

[0032] Taking the user-segmented insurance application process version dynamic diversion method in this embodiment as an example, please refer to Figure 1. Figure 1 is a schematic diagram of the hardware environment of an optional user-segmented insurance application process version dynamic diversion method provided in this application embodiment. As shown in Figure 1, the hardware environment of the user-segmented insurance application process version dynamic diversion method includes: a terminal 102, and a server 104 connected to the terminal 102 via a network. The server 104 is used to deploy and execute the relevant logic of the user-segmented insurance application process version dynamic diversion method, including the functions of modules such as data acquisition, user segmentation, version allocation, data collection and statistics, and strategy decision-making, so as to realize the dynamic optimization and diversion of the insurance application process version; the terminal 102 is used to send the user's insurance application request to the server and display the corresponding insurance application process version page allocated by the server, while collecting and feeding back the user's real-time behavior data.

[0033] The dynamic diversion method for insurance application process versions based on user segmentation in this embodiment can be applied to scenarios such as insurance application process version optimization in the field of insurance technology, such as: online insurance sales platforms, mobile insurance applications, and insurance application process version management on the official website of insurance companies. In this embodiment, the above-mentioned dynamic diversion method for insurance application process versions based on user segmentation is illustrated by taking the optimization of the insurance application process version of an online insurance sales platform as an example.

[0034] The existing technology has fundamental problems in the optimization of insurance application process versions, mainly manifested in low experimental efficiency and response delay. This is mainly because traditional static A / B testing must wait for a complete experimental cycle to draw conclusions, and cannot release high-conversion-rate version traffic in real time, resulting in missed promotion opportunities and potential loss of sales; at the same time, the existing technology has poor dynamic adaptability and is difficult to capture changes in users' real-time behavioral intentions, resulting in a mismatch between recommended paths and user needs; in addition, the business targeting is weak.The failure to deeply integrate with the characteristics of insurance business, the lack of verification mechanisms for regional compliance requirements and differentiated value of insurance products, and the existence of insufficient business adaptation and compliance risks.

[0035] To solve the above problems, this embodiment provides a method for dynamic diversion of insurance application process versions based on user segmentation running on the above-mentioned server. Please refer to Figure 2. Figure 2 is a flowchart of an optional method for dynamic diversion of insurance application process versions based on user segmentation provided in this application embodiment. As shown in Figure 2, the method for dynamic diversion of insurance application process versions based on user segmentation in this application embodiment specifically includes the following steps: Step S201: In response to the user's insurance application request, obtain user attribute data and real-time behavior data; Step S202: According to the user attribute data and real-time behavior data, divide the user into a predefined user segment; Step S203: Based on the user segment and the current effective user diversion ratio, assign the corresponding insurance application process version to the user; Step S204: Collect and statistically analyze the conversion rate data of each insurance application process version in real time, and determine whether the preset dynamic adjustment conditions are met; Step S205: If it is determined that the dynamic adjustment conditions are met, calculate and update the user diversion ratio according to the conversion rate data, and assign the insurance application process version to subsequent users according to the updated user diversion ratio.

[0036] Through the above steps S201 to S205, by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio, the inherent defects of traditional static A / B testing, such as long experimental cycles and response delays, are effectively overcome. This allows for timely amplification of traffic for high-conversion-rate versions, thereby significantly improving experimental efficiency and overall conversion rate. At the same time, by combining the user segmentation mechanism with the dynamic diversion strategy, personalized process recommendations based on user attributes and real-time behavioral intentions are realized, enhancing the system's adaptability to dynamic user needs. This also provides a technical foundation for the deep integration of insurance business rules (such as regional compliance verification and insurance type differentiation weights), thereby ensuring the accuracy and compliance of the business process while pursuing maximum conversion efficiency.

[0037] The following explanation, in conjunction with Figure 2, illustrates the dynamic diversion method for the insurance application process version based on user segmentation in this embodiment.

[0038] In the technical solution of step S201, in response to the user's insurance application request, user attribute data and real-time behavior data are obtained, specifically including: when it is detected that the user triggers a specific insurance application entry interaction event on the insurance product details page, the system generates an insurance application request and establishes an independent user session; the basic attribute information of the logged-in user is obtained by calling the user center service interface, and the user's interaction behavior sequence in the current session is captured in real time through front-end tracking technology; the obtained structured attribute data and unstructured behavior event data are associated with the same session ID, and encapsulated into a unified format context data packet, which is then transmitted to the downstream cluster decision module.

[0039] In this embodiment, a user's insurance application request refers to an operation instruction initiated by the user expressing their intention to purchase an insurance product. This is typically manifested as the user clicking the "Apply Now" button, submitting an application form, or issuing an application instruction via a voice assistant. This request is a structured event containing metadata such as timestamps, user ID, device fingerprint, and insurance type ID. For example, if a user selects a "Three-Year Car Insurance Plan" in a mobile app and clicks the "Next" button, the system will generate an insurance application request containing information such as {user_id: "U12345", product_id: "P001", timestamp: "2024-06-15T10:30:00Z"}, send it to the server, and establish an independent session.

[0040] User attribute data refers to static or long-term stable data related to the user's identity, preferences, and risk characteristics, typically accumulated during user registration or historical interactions. This includes, but is not limited to, geographical location, age, gender, occupation, historical insurance application records, credit score, etc. For example: Geographic location: The "permanent residence" filled in by the user during registration is "City A, District B"; Age: The ID card information associated with the user account shows "35 years old"; Historical insurance records: The system database stores records of the user's purchase of "travel accident insurance" and "home property insurance" in the past 2 years.

[0041] Real-time behavioral data: refers to the dynamic operation data generated by the user in the current insurance process version, reflecting their real-time intentions and interaction status. It includes page browsing path, dwell time, form filling progress, button click frequency, etc. For example: Page dwell time: The user stayed on the "Health Declaration" page for 2 minutes and 30 seconds; Form filling progress: The user has filled in the "Vehicle Owner Information" and "Vehicle Information" modules in the car insurance application form, but has not yet completed the "Policyholder Signature"; Click operation: The user clicked the "Premium Calculation" button repeatedly 3 times, and finally clicked the "Submit" button.

[0042] As an optional embodiment, the specific implementation process for obtaining user attribute data and real-time behavior data in response to a user's insurance application request is as follows: When a user accesses the insurance sales platform and initiates an insurance application operation through a terminal (such as a mobile phone or computer), the system captures the user's insurance application request through the front-end interface.

[0043] At this time, the server-side data acquisition module is started synchronously. First, it extracts pre-stored user attribute data (such as geographical location, age, historical insurance records, etc.) from the request header or user session. At the same time, it collects user behavior data in the insurance application process version in real time through the tracking technology (such as page dwell time, form filling progress, click operations, etc.).

[0044] For example, if a user accesses the car insurance application page in region A, the system will record its geographical location as "A" and track its real-time progress when filling in the vehicle information form (such as having completed 50% of the required fields).

[0045] These data are encrypted and then transmitted to the server, providing a foundation for subsequent user grouping and dynamic traffic allocation. It should be noted that in this embodiment, the specific steps for collecting user behavior data in the insurance application process using the data collection technology, as well as the specific steps for encrypting the acquired data, are not specifically limited. Please refer to the relevant content in the prior art.

[0046] Through the above technical means, by combining static attributes with dynamic intent signals through multi-source data fusion technology, a panoramic user profile is provided for subsequent grouping decisions, effectively solving the problem of insufficient accuracy of path recommendation caused by the single data dimension in traditional solutions.

[0047] In the technical solution of step S202, users are divided into predefined user groups based on user attribute data and real-time behavior data.

[0048] In this embodiment, user grouping is achieved by combining user attribute data and real-time behavior data to divide users into predefined groups with similar characteristics or needs, so as to achieve accurate process allocation and dynamic adaptation. Through multi-dimensional data fusion analysis, refined classification of user groups is achieved, providing a foundation for differentiated services. The system categorizes users into four core groups based on geographical location risk level and real-time user intent intensity: high-risk high intent, high-risk, high intent, and ordinary intent.

[0049] For example: High-risk area user grouping: If a user's geographical location matches a preset list of high-risk areas (such as areas with strict insurance regulation), they are further divided into two categories based on their form completion progress: high-risk high intent (completion progress > first preset threshold) and high-risk (completion progress ≤ first preset threshold). Ordinary area user grouping: If a user is not in a high-risk area, they are divided into two categories based on page dwell time and form completion progress: high intent (dwell time > second preset threshold and completion progress > third preset threshold) and ordinary intent (dwell time ≤ second preset threshold or completion progress ≤ third preset threshold).

[0050] As an optional embodiment, users are divided into predefined user groups based on user attribute data and real-time behavior data, including: obtaining the geographic location attribute in the user attribute data and matching the geographic location attribute with a preset high-risk area list, wherein the high-risk area list is pre-configured according to the strictness of insurance regulatory policies in each region; if the match is successful, the user is determined to be in a preset high-risk area, and the first grouping logic is executed: obtaining the form filling progress in the real-time behavior data, wherein the form filling progress is obtained by tracking the ratio of the number of required form fields completed by the user in the insurance application process version to the total number of required form fields; determining whether the form filling progress is greater than a first preset threshold; if the form filling progress is determined to be greater than the first preset threshold, the user is divided into a user group labeled as high-risk and high-intention; if the form filling progress is determined to be less than or equal to the first preset threshold, the user is divided into a user group labeled as high-risk.If the matching fails, it is determined that the user is not in a high-risk area, and the second grouping logic is executed: obtain the page dwell time and form filling progress from the real-time behavior data; determine whether the page dwell time is greater than the second preset threshold, and at the same time determine whether the form filling progress is greater than the third preset threshold; if it is determined that the page dwell time is greater than the second preset threshold and the form filling progress is greater than the third preset threshold, then the user is divided into the user group labeled as high intention; if it is determined that the page dwell time is less than or equal to the second preset threshold and / or the form filling progress is less than or equal to the third preset threshold, then the user is divided into the user group labeled as ordinary intention.

[0051] In this embodiment, the high-risk area list is a set of high-risk area codes pre-set based on regulatory requirements. It is determined according to the regional risk rating issued by the insurance regulatory agency and is updated according to a certain period, such as monthly dynamic updates.

[0052] The label can also be called the group label, which is used to characterize the identifier of the group type to which the user belongs. Specific label description: High-risk high intention group: refers to the user group that is in a high-risk area and shows a strong intention to purchase insurance. High-risk group: refers to the user group in high-risk areas but whose intention to purchase insurance is not yet clear. High-intention group: refers to the user group in non-high-risk areas who show a clear intention to purchase insurance. Ordinary intention group: refers to the user group in non-high-risk areas with a moderate degree of intention.

[0053] The form filling progress is the percentage of required fields that the user has completed in the current version of the insurance process out of the total required fields. It can be calculated by the following formula: Form filling progress = (number of required fields filled / total number of required fields) × 100%.

[0054] Through the above grouping mechanism, the flexible configuration of the threshold parameter enables the grouping strategy to ensure business compliance and respond to changes in user intention in real time, effectively solving the problems of single user profile and rigid strategy in traditional solutions. The accurate matching of each group label with the subsequent process version provides a scientific decision basis for dynamic diversion.

[0055] In this embodiment, the grouping logic is a rule that divides users into specific groups based on the combination of user attributes and real-time behavior. The grouping logic includes two types: high-risk area grouping logic and ordinary area grouping logic.

[0056] High-risk area grouping logic: If the high-risk area list is successfully matched, the first grouping logic is executed, that is, if the form filling progress is greater than the first preset threshold, it is classified as high-risk and high-intention; otherwise, it is classified as high-risk.

[0057] Ordinary area grouping logic: If the high-risk area list is not matched, the second grouping logic is executed, that is, if the page dwell time is greater than the second preset threshold and the form filling progress is greater than the third preset threshold, it is classified as high-intention; otherwise, it is classified as ordinary-intention.

[0058] It can be understood that the first preset threshold is used to judge the form completion rate of users in high-risk areas. Optionally, the first preset threshold can be 50%; the second preset threshold is used to judge the page dwell time. Optionally, the second preset threshold can be 50%.60 seconds; the third preset threshold is used to judge the form completion rate of users in non-high-risk areas. Optionally, the third preset threshold can be the same as the first preset threshold or different from the first preset threshold, for example, the third preset threshold is 30%.

[0059] According to the above technical means, by deeply integrating user static attributes such as geographical location with dynamic behavioral intentions such as form filling progress and page dwell time, a refined real-time grouping model is constructed, which can automatically identify and distinguish user groups with different characteristics such as "high compliance requirements" and "high conversion potential". On the one hand, this mechanism can ensure that users in high-risk areas are accurately guided to the insurance process version containing mandatory compliance steps, avoiding regulatory risks from the source. On the other hand, it can prioritize the allocation of high-intent users to the fast channel with better experience, effectively improving the success rate of the insurance process. Thus, while ensuring the compliance of the business, it achieves the simultaneous optimization of traffic allocation efficiency and accuracy.

[0060] In the technical solution of step S203, the corresponding insurance process version is allocated to the user based on the user group and the current effective user diversion ratio.

[0061] In this embodiment, the currently effective user traffic allocation ratio refers to the real-time user traffic ratio configuration allocated by the system to each insurance application process version at a specific point in time. It is periodically adjusted based on real-time conversion rate data, such as updating every 5 minutes. Each ratio value corresponds to a specific process version identifier, and the sum of the traffic allocation ratios of each version is constant at 100%.

[0062] The insurance application process refers to a set of orderly operation steps designed for users to complete insurance purchases. There are one or more different process implementation schemes under the same insurance product, that is, different insurance application process versions under the same insurance product. Specifically: each version has differences in interface design, number of steps, interaction methods, etc. Based on specific practices, one or more insurance application process versions are pre-assigned to each user group.

[0063] For example: the simplified quick version (version A) includes 3 core steps, collapsed non-mandatory fields, and default recommended options, which is suitable for high-intent users and repeat purchase users. The standard guided version (version B) includes 5 standard steps, complete form display, appropriate guidance prompts, etc., which is suitable for ordinary intent users and new users. The complete compliant version (Version C) includes 7 complete steps, including dual recording, mandatory reading time, etc., and is suitable for users in high-risk areas and areas with strict regulatory requirements.

[0064] As an optional embodiment, based on user groups and the currently effective user diversion ratio, a corresponding insurance process version is assigned to the user, including: determining one or more candidate insurance process versions according to the group tags corresponding to the user groups; obtaining the currently effective user diversion ratio, wherein the user diversion ratio defines the proportion of user traffic that each candidate version should be allocated at the current moment; and using a consistent hash algorithm to identify the user according to the currently effective user diversion ratio.The user is divided into groups to assign a unique target insurance application process version from the candidate application process versions.

[0065] In this embodiment, the group label is used to identify the group type to which the user belongs. Specific label descriptions: High-risk high-intention group: refers to a group of users in high-risk areas who show a strong intention to apply for insurance. High-risk group: refers to a group of users in high-risk areas whose intention to apply for insurance is not yet clear. High-intention group: refers to a group of users in non-high-risk areas who show a clear intention to apply for insurance. Ordinary intention group: refers to a group of users in non-high-risk areas with a moderate level of intention.

[0066] The candidate application process version refers to the set of application process versions that are pre-selected based on the user group label and that meet the business rules for displaying to users in that group.

[0067] The candidate insurance application process version selection logic may include: the candidate insurance application process version corresponding to high-risk, high-intention / high-risk groups is the complete compliant version; the candidate insurance application process version for high-intention groups is the simplified quick version and the standard guided version; the candidate insurance application process version for ordinary intention groups is the standard guided version.

[0068] The target insurance application process version refers to the specific insurance application process version that is ultimately selected from the candidate version set through a consistent hashing algorithm and is actually displayed to the user. It is based on the candidate version set and combined with the real-time traffic distribution ratio for precise traffic allocation. For example: the candidate set for a high-intention user is {version A, version B}, and version A is finally selected after hashing and bucketing.

[0069] It should be noted that for users in high-risk areas, due to business restrictions, the candidate insurance application process versions only include the compliant version, so the target version is forced to be the compliant version; for ordinary users, the candidate insurance application process version set contains multiple versions, and the target insurance application process version is dynamically determined by the algorithm. The specific steps for determining the candidate insurance application process version through consistent hashing and bucketing in this embodiment are not specifically limited, please refer to the relevant content of the prior art.

[0070] Based on the above technical means, by combining business-oriented user grouping rules with data-driven traffic diversion ratio strategies, while ensuring that users with high compliance requirements are accurately guided to the safe process, the uniformity, stability, and consistency of user sessions are achieved by using a consistent hashing algorithm. This mechanism not only ensures that the traffic diversion strategy strictly follows the compliance constraints of the business specification page 9 / 16 13 CN 121504635 A, avoiding regulatory risks caused by version misallocation, but also ensures the accuracy of experimental data and system load balancing in A / B testing through predictable and low-conflict hash bucketing technology, thereby maintaining high performance and high reliability of traffic diversion decisions even under complex business rules.

[0071] In the technical solution of step S204, the conversion rate data of each insurance application process version is collected and statistically analyzed in real time, and it is determined whether the preset dynamic adjustment conditions are met.

[0072] In this embodiment, the conversion rate data refers to the set of quantitative indicators for measuring the conversion effect of each insurance application process version.Specifically, the conversion rate data includes a multi-dimensional set of indicators: basic statistical indicators such as the number of users accessing the insurance application process (the number of unique users accessing the insurance application process version within the statistical window, for example, 1,200 people), the number of successful insurance applications (the number of users who completed the insurance application and successfully paid, for example, 102 people), and the original conversion rate (calculated as the number of successful insurance applications / the number of users accessing the application process × 100%, for example, 8.5%); time dimension data includes the statistical window identifier (such as the start and end time of the 1-hour scrolling window) and the data generation timestamp; in addition, it also includes version-related metadata (such as the insurance application process version ID and insurance type). This structured data is generated in real time through a stream processing engine, providing comprehensive and reliable data support for dynamic traffic diversion decisions.

[0073] The dynamic adjustment condition refers to the preset judgment criteria for triggering the recalculation of the traffic diversion ratio. Specifically, it refers to the time difference between the current time and the time point of the last successful execution of dynamic adjustment, which reaches or exceeds the preset adjustment period.

[0074] As an optional embodiment, determining whether the preset dynamic adjustment conditions are met includes: obtaining the current system time and retrieving the preset adjustment period, wherein the adjustment period defines the time interval for dynamic adjustment; calculating the time difference between the current time and the time point of the last successful dynamic adjustment; determining whether the time difference reaches or exceeds the adjustment period; if the time difference reaches or exceeds the adjustment period, it is determined that the dynamic adjustment conditions are met.

[0075] In this embodiment, the adjustment period refers to the fixed time interval preset by the system that triggers the recalculation of the dynamic diversion ratio. Its setting is the result of a trade-off between business needs, data science, and system performance. If the period is too short (e.g., 1 minute), the strategy may be adjusted based on unstable random fluctuations; if the period is too long (e.g., 1 hour), the effect of "real-time" optimization cannot be achieved. In the insurance application scenario, it usually takes several minutes for a user to go from entering the process to completing the payment. A moderate time window helps to obtain more stable conversion rate data. Moreover, frequent full calculations will put pressure on data processing and decision engines, so it is necessary to set an interval that is friendly to system resources. In a preferred embodiment, the adjustment period is set to 5 minutes. Within this period, it is possible to capture significant increases in conversion rates due to marketing activities, time period changes, or the attractiveness of new versions, thereby rapidly amplifying traffic from high-value versions; and to accumulate a sufficient number of user visits and order samples, so that the calculated conversion rate has certain statistical significance, avoiding frequent strategy fluctuations due to the randomness of individual orders.

[0076] According to the above technical means, by introducing an automated triggering mechanism based on a fixed time period, the dynamic adjustment is transformed from an event-driven mode that relies on manual intervention or data fluctuations into a systematic time-driven mode, realizing the periodic and automated self-optimization of the traffic diversion strategy; this not only completely avoids the need for traditional A / B testing to wait for the complete experimental period to end before...The response delay issue that leads to conclusions enables the system to respond quickly to changes in conversion rate at a preset pace and amplify high-potential version traffic in a timely manner, thereby significantly improving experimental efficiency and resource utilization. At the same time, it ensures the controllability, predictability, and stability of the system's adjustment behavior, avoiding frequent strategy oscillations caused by instantaneous fluctuations in real-time data.

[0077] In the technical solution of step S205, if it is determined that the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated based on the conversion rate data, and the insurance process version is allocated to subsequent users according to the updated user diversion ratio.

[0078] As an optional embodiment, calculating and updating the user diversion ratio based on conversion rate data includes: obtaining the real-time conversion rate of each insurance application process version within the current statistical window, wherein the real-time conversion rate is the ratio of the number of users who successfully completed the insurance application process version to the total number of users accessing the insurance application process version; determining the insurance type weight based on the type of insurance product carried by each insurance application process version; determining the weighted conversion rate corresponding to each insurance application process version based on the real-time conversion rate and insurance type weight of each insurance application process version through the weighted conversion rate calculation formula; and determining the user diversion ratio corresponding to each insurance application process version through the diversion ratio formula based on the weighted conversion rate and the sum of the weighted conversion rates of all insurance application process versions.

[0079] In this embodiment, the real-time conversion rate refers to the instantaneous conversion effect indicator of a single insurance application process version within a specific statistical window. In practice, the real-time conversion rate is calculated based on the rolling time window (e.g., the past hour) and the real-time conversion rate formula. For the i-th insurance application process version, the real-time conversion rate formula is: CRi = (Ni / Sumi) × 100%, where Ni is the number of users who successfully completed the insurance application, Sumi is the total number of users who visited, and CRi is the real-time conversion rate.

[0080] For example: During the period from 10:00 to 11:00, version A had a total of 1,200 users who visited and 102 users who successfully applied for insurance. Therefore, the real-time conversion rate of version A is 8.5%.

[0081] The insurance type weight is a business value coefficient set based on the type of insurance product, used to adjust the importance of different insurance types in the triage decision. In practice, each insurance product will be pre-configured with a basic insurance type weight. For example: the default weight of accident insurance is 1.2; the default weight of million-dollar medical insurance is 1.0; and the default weight of life insurance is 1.5. The insurance type weight can also be dynamically adjusted according to the actual situation. For example: During high-value periods (such as evening), the weight of accident insurance is automatically increased to 1.5.

[0082] The weighted conversion rate is a comprehensive evaluation indicator after adjusting the real-time conversion rate according to the insurance type weight. For the i-th insurance application process version, the weighted conversion rate calculation formula is: WCi=CRi×Wi, where WCi is the weighted conversion rate and CRi is the real-time conversion rate.The rate, where Wi is the insurance type weight. For example, if the real-time conversion rate of version A is 8.5% and its insurance type weight is 1.2, then its weighted conversion rate is calculated to be 10.2% using the above weighted conversion rate formula.

[0083] The diversion ratio formula is the core algorithm used to normalize the weighted conversion rate into a probability distribution.

[0084] For the i-th insurance application process version, the diversion ratio calculation formula is: Pi = WCi / SumWC, where WCi is the weighted conversion rate and SumWC is the sum of the weighted conversion rates of all versions.

[0085] In specific practice, for the i-th insurance application process version, its weighted conversion rate is calculated according to the weighted conversion rate formula WCi=CRi×Wi, where WCi is the weighted conversion rate, CRi is the real-time conversion rate, and Wi is the insurance type weight; the sum of the weighted conversion rates of all insurance application process versions is calculated, i.e., SumWC= Σ (WCi); for the i-th insurance application process version, its target diversion ratio is calculated according to the formula Pi=WCi / SumWC.

[0086] Based on the above technical means, by introducing the insurance type weight coefficient and weighting its real-time conversion rate, a diversion ratio decision model that deeply integrates business value orientation is constructed, so that the diversion strategy not only responds to the conversion efficiency of each process version, but also accurately reflects the differentiated business value of different insurance products; this mechanism automatically maps the weighted conversion rate to the diversion ratio through the normalization algorithm, realizing the optimal allocation of traffic resources to the "high conversion + high value" combination, thereby ensuring the mathematical rationality and business adaptability of the diversion decision.

[0087] In the technical solution of step S205, if it is determined that the dynamic adjustment conditions are not met, the currently effective user diversion ratio will continue to be used to allocate insurance process versions to subsequent users.

[0088] In specific practice, all calculated target diversion ratios are subject to constraint adjustment: it is determined one by one whether the target diversion ratio of each version is lower than the minimum protection ratio of 5%; if there is a version with a target diversion ratio lower than 5%, its diversion ratio is increased to 5%, and the diversion ratios of the remaining versions are reduced accordingly according to the original calculated ratio to ensure that the sum of all diversion ratios is 100%; the finally adjusted diversion ratio is updated to the currently effective user diversion ratio.

[0089] As an optional embodiment, the insurance product weight is determined according to the insurance product type carried by each insurance application process version, including: obtaining the basic weight corresponding to the insurance product type according to the insurance product type and a predefined insurance product weight configuration table; obtaining the current system time and determining whether the current time is within a preset target period, wherein the target period includes holidays, large-scale promotional activities, and high-value periods of the day; if it is determined that the current time is within the preset target period...During the period, the final effective insurance type weight is determined based on the current time and the preset time period-weight mapping table. The time period-weight mapping table stores the mapping relationship between a specific time period and the corresponding insurance type weight in the form of key-value pairs. If it is determined that the current time is not within the preset target period, the basic weight is used as the final effective insurance type weight.

[0090] In this embodiment, the insurance type weight configuration table is a static parameter table predefined by the system. It stores the benchmark weight coefficients corresponding to different insurance product types in a structured form. Its data comes from a comprehensive evaluation of product profit margin, strategic priority and long-term value. For example, accident insurance is configured as 1.2 and life insurance is configured as 1.5, providing an initial basis for weight calculation.

[0091] The basic weight refers to the fixed weight value used by the insurance type under normal conditions. It serves as a benchmark anchor point for dynamic adjustment and is not affected by time factors. For example, accident insurance always uses 1.2 as the basic weight, ensuring the stability of the weight system in the default scenario.

[0092] The target period is a pre-defined set of special time periods, including statutory holidays, large-scale marketing campaigns, and daily high-value periods (such as 20:00-22:00). During these periods, the weight adjustment mechanism is triggered due to user activity or business strategy needs.

[0093] The time period-weight mapping table uses a key-value pair structure to store the mapping relationship between time conditions and weight coefficients. For example, the key "20:00-22:00" corresponds to the value {"Accident Insurance": 1.5}, enabling the system to quickly obtain dynamic adjustment parameters through time matching.

[0094] The final effective insurance type weight is the actual usage weight after time dimension calibration. For example, the weight of accident insurance increases from 1.2 to 1.8 during the evening period, and finally participates in the traffic diversion ratio calculation, realizing the accurate adaptation of business strategy and time environment.

[0095] Based on the above technical means, by introducing a dynamic coupling mechanism between the time dimension and business strategy, the static basic weight of insurance types is intelligently associated with preset time factors such as high-value periods and promotional activity periods, realizing adaptive optimization of the diversion strategy in different time contexts; this mechanism can automatically identify the key business period and increase the weight priority of high-value insurance types, ensuring that traffic resources are tilted towards high-yield products during the golden period, thereby significantly improving the overall input-output ratio and resource utilization of marketing activities without manual intervention in the system, while enhancing the system's dynamic response capability to changes in the market environment and the accuracy of strategy execution.

[0096] As an optional embodiment, before allocating insurance process versions to subsequent users according to the updated user diversion ratio, the method further includes: after calculating the preliminary user diversion ratio adjustment plan based on conversion rate data, identifying one or more target insurance process versions that plan to increase the diversion ratio from the plan; for each target insurance process version that plans to increase the diversion ratio, executing a real-time compliance verification process, specifically including: obtaining the user ID of the current user to be diverted, and according toThe user ID queries the corresponding user attribute data, which includes at least geographical location and age. Based on the user's geographical location, age, and the identifier of the target insurance process version as input, the compliance verification service is called. The system queries the pre-set compliance rule library to verify whether the target insurance process version meets all mandatory compliance requirements for the corresponding user's location, insurance type, and user attributes. If the verification passes, the target insurance process version meets the compliance requirements, and the diversion ratio increase operation is allowed. If the verification fails, the target insurance process version does not meet the compliance requirements, and the diversion ratio increase operation for the target insurance process version is canceled.

[0097] In this embodiment, compliance verification refers to verifying whether the target insurance process version meets the mandatory compliance requirements for the user's location, insurance type, and attributes before diversion. For example, if user C (age 17) is assigned to the high-intention group, the system verification finds that his age does not meet the insurance requirements for a certain children's insurance, and the diversion ratio increase for that process is canceled. If user D (located in a high-risk area) is recommended to the simplified insurance process version, but the compliance rules require that the area must complete facial recognition, the system blocks this recommendation. Instruction manual, pages 12 / 16, CN 121504635 A

[0098] In this embodiment, the rule base stores at least the following mandatory compliance rules: Regional rules: define the mandatory requirements for the insurance application process in different administrative regions, such as requiring a "dual recording" function in a designated region; Insurance type rules: define the display and operation requirements for specific insurance product types, such as requiring a mandatory reading time for the full text of the terms and conditions for long-term life insurance; User rules: define restrictive conditions based on user attributes, such as prohibiting the recommendation of high-risk investment-linked insurance products to users over 60 years of age.

[0099] Based on the above technical means, by embedding a real-time compliance verification step before the execution of the diversion strategy, a business rule-driven security protection mechanism is constructed. This mechanism can automatically identify and block diversion operations that violate regional regulatory policies or user-specific restrictions based on user location, age, and target process characteristics. This mechanism upgrades compliance verification from offline checks to pre-emptive real-time blocking, effectively avoiding the risk of regulatory penalties caused by misallocation of process versions. Furthermore, the system's automated judgment ensures agile optimization of the diversion strategy within the business compliance framework, achieving a balance between business security and experimental efficiency.

[0100] It should be noted that, for the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations. However, those skilled in the art should understand that this application is not limited by the described order of actions, because according to this application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0101] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM (Read-Only Memory) / RAM (Random Access Memory), magnetic disk, optical disk), including several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0102] According to another aspect of the embodiments of this application, a user-segmented insurance process version dynamic diversion system is also provided for implementing the above-mentioned user-segmented insurance process version dynamic diversion method. Please refer to Figure 3. Figure 3 is a structural block diagram of an optional user-segment-based dynamic diversion system for insurance application process versions provided in an embodiment of this application. As shown in Figure 3, the system 300 may include: a data acquisition module 301, used to acquire user attribute data and real-time behavior data in response to a user's insurance application request; a user segmentation module 302, used to divide users into predefined user segments according to user attribute data and real-time behavior data; a version allocation module 303, used to allocate corresponding insurance application process versions to users based on user segments and the currently effective user diversion ratio; a data collection and statistics module 304, used to collect and statistically analyze the conversion rate data of each insurance application process version in real time; and a strategy decision module 305, used to determine whether preset dynamic adjustment conditions are met; when it is determined that the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated based on the conversion rate data.

[0103] It should be noted that the data acquisition module 301 in this embodiment can be used to execute the above step S201, the user grouping module 302 in this embodiment can be used to execute the above step S202, the version allocation module 303 in this embodiment can be used to execute the above step S203, the data collection and statistics module 304 in this embodiment can be used to execute the above step S204, and the strategy decision module 305 in this embodiment can be used to execute the above step S205. The version allocation module 303 is also used to allocate the insurance application process version to subsequent users according to the updated user diversion ratio.

[0104] Regarding the dynamic diversion system for insurance application process versions based on user grouping in this embodiment, its data acquisition module 301, user grouping module 302, version allocation module 303, data collection and statistics module 304, and strategy decision module 305The specific method for implementing the above-mentioned dynamic diversion method for the insurance application process based on user segmentation has been described in detail in the embodiments of the dynamic diversion method for the insurance application process based on user segmentation, and will not be elaborated here.

[0105] It is understood that the technical solution provided in this embodiment, in the dynamic diversion system for the insurance application process based on user segmentation, each module effectively overcomes the inherent defects of long experimental cycles and response delays in traditional static A / B testing by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio. It can promptly amplify the traffic of high conversion rate versions, thereby significantly improving experimental efficiency and overall order conversion rate. At the same time, by combining the user segmentation mechanism with the dynamic diversion strategy, it not only realizes personalized process recommendation based on user attributes and real-time behavioral intentions, enhancing the system's adaptability to dynamic user needs, but also provides a technical foundation for the deep integration of insurance business rules (such as regional compliance verification and insurance type differentiation weights), thereby ensuring the accuracy and compliance of the business process while pursuing maximum conversion efficiency.

[0106] In addition to the modules described above, the apparatus in this embodiment may also include a module that executes any method in any of the embodiments of the dynamic diversion method for the insurance application process version based on user segmentation.

[0107] It should be noted here that the examples and application scenarios implemented by the modules and corresponding steps described above are the same, but are not limited to the content disclosed in the above embodiments. It should be noted that the modules described above, as part of the apparatus, can run in a hardware environment that implements the method shown in FIG1, and can be implemented by software or by hardware, wherein the hardware environment includes a network environment.

[0108] According to another aspect of the embodiments of this application, an electronic device for implementing the above-described dynamic diversion method for the insurance application process version based on user segmentation is also provided. The electronic device may be a server, a terminal, or a combination thereof.

[0109] According to another embodiment of this application, an electronic device is also provided. Please refer to FIG4. FIG4 is a structural block diagram of an optional electronic device provided in an embodiment of this application. As shown in FIG4, the electronic device may include: a processor 1501, a communication interface 1502, a memory 1503, and a communication bus 1504, wherein the processor 1501, the communication interface 1502, and the memory 1503 communicate with each other through the communication bus 1504.

[0110] The memory 1503 is used to store computer programs; the processor 1501 is used to execute the program stored in the memory 1503 to implement the following steps: Step S201, in response to the user's insurance application request, obtain user attribute data and real-time behavior data; Step S202, according to the user attribute data and real-time behavior data, divide the user into a predefined user group; Step S203, based on the user group and the currently effective user diversion ratio, assign the corresponding insurance application process version to the user;Step S204: Collect and statistically analyze the conversion rate data of each insurance application process version in real time, and determine whether the preset dynamic adjustment conditions are met; Step S205: If the dynamic adjustment conditions are met, calculate and update the user diversion ratio based on the conversion rate data, and allocate the insurance application process version to subsequent users according to the updated user diversion ratio.

[0111] It is understood that the technical solution provided in this embodiment, the processor of the electronic device, by collecting conversion rate data in real time and dynamically adjusting the user diversion ratio, effectively overcomes the inherent defects of traditional static A / B testing, such as long experimental cycles and response delays. It can promptly amplify the traffic of high conversion rate versions, thereby significantly improving experimental efficiency and overall order conversion rate. At the same time, by combining the user grouping mechanism with the dynamic diversion strategy, it not only realizes personalized process recommendation based on user attributes and real-time behavioral intentions, enhancing the system's adaptability to dynamic user needs, but also provides a technical foundation for the deep integration of insurance business rules (such as regional compliance verification and insurance type differentiation weights), thereby ensuring the accuracy and compliance of business processes while pursuing maximum conversion efficiency.

[0112] Optionally, in this embodiment, the above-mentioned communication bus can be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The communication bus can be divided into address bus, data bus, control bus, etc. For ease of representation, only one thick line is used in the figure, but this does not mean that there is only one bus or one type of bus. The communication interface is used for communication between the above-mentioned electronic device and other devices.

[0113] The memory may include random access memory (RAM), or non-volatile memory (NVM), such as at least one disk storage device. Optionally, the memory may also be at least one storage device located away from the aforementioned processor.

[0114] The above-mentioned processor may be a general-purpose processor, which may include but is not limited to: CPU (Central Processing Unit), NP (Network Processor), etc.; it may also be DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array).Array (Field Programmable Gate Array) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0115] This application embodiment also provides a computer-readable storage medium, the storage medium including a stored program, wherein the program executes the method steps of the above method embodiments when it runs.

[0116] Optionally, in this embodiment, the above storage medium may include, but is not limited to: USB flash drive, ROM, RAM, mobile hard disk, magnetic disk or optical disk and other media that can store program code.

[0117] The sequence numbers of the above embodiments of this application are only for description and do not represent the superiority or inferiority of the embodiments.

[0118] If the integrated unit in the above embodiments is implemented in the form of software functional unit and sold or used as an independent product, it can be stored in the above computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in the storage medium and includes several instructions to cause one or more computer devices (which may be personal computers, servers or network devices, etc.) to execute all or part of the steps of the methods of the various embodiments of this application.

[0119] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed client can be implemented in other ways. Among them, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical functional division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of units or modules may be electrical or other forms.

[0121] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Pages 15 / 16 of the specification 19 CN 121504635 A. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution provided in this embodiment.

[0122] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0123] The above are merely preferred embodiments of this application. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of this application, and these improvements and modifications should also be considered within the scope of protection of this application. Instruction sheet 16 / 16 page 20 CN 121504635 A Figure 1 Figure 2 Instruction drawing 1 / 2 page 21 CN 121504635 A Figure 3 Figure 4 Instruction sheet drawing 2 / 2 page 22 CN 121504635 A {HARTEND-26018-HKSPT / 02504766v1} Abstract This application pertains to the technical field of insurance underwriting business, specifically to a dynamic diversion method and system for insurance underwriting process versions based on user segmentation. By collecting conversion rate data in real time and dynamically adjusting the user diversion ratio, this application overcomes the shortcomings of traditional static A / B testing experiments, such as long cycles and delayed responses. It can promptly amplify the traffic of high-conversion-rate versions, significantly improving the efficiency of experiments and the success rate of insurance underwriting. Meanwhile, by combining user segmentation with dynamic diversion ratios for process allocation, it not only achieves personalizedrecommendations based on user attributes and real-time intentions, enhancing dynamic adaptability, but also provides a technical foundation for the deep integration of insurance business rules through the segmentation mechanism. Thus, while improving efficiency, it ensures the accuracy and compliance of business operations.

Claims

1. A method for dynamic traffic diversion in the insurance application process based on user segmentation, characterized in that, include: In response to a user's insurance application request, obtain user attribute data and real-time behavioral data; Based on the user attribute data and the real-time behavior data, the users are divided into predefined user groups; Based on the user grouping and the currently effective user diversion ratio, assign the corresponding insurance application process version to the user; Real-time collection and statistics of conversion rate data for each insurance application process version, and determination of whether the preset dynamic adjustment conditions are met; If the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated based on the conversion rate data, and the insurance process version is assigned to subsequent users according to the updated user diversion ratio.

2. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, Based on the user attribute data and the real-time behavior data, the users are divided into predefined user groups, including: Obtain the geographic location attribute from the user attribute data, and match the geographic location attribute with a preset list of high-risk areas, wherein the list of high-risk areas is pre-configured based on the strictness of insurance regulatory policies in each region; If a match is successful, the user is determined to be in a preset high-risk area, and the first grouping logic is executed: obtain the form filling progress in the real-time behavior data, wherein the form filling progress is obtained by tracking the ratio of the number of required form fields completed by the user in the insurance process version to the total number of required form fields; determine whether the form filling progress is greater than a first preset threshold. If it is determined that the form completion progress is greater than the first preset threshold, then the user will be classified into a user group labeled as high-risk and high-intention. If it is determined that the form completion progress is less than or equal to the first preset threshold, then the user will be classified into a user group labeled as high-risk. If the matching fails, it is determined that the user is not in a high-risk area, and the second grouping logic is executed: obtain the page dwell time and form filling progress from the real-time behavior data; determine whether the page dwell time is greater than a second preset threshold, and at the same time determine whether the form filling progress is greater than a third preset threshold; If it is determined that the duration of the page stay is greater than the second preset threshold and the form filling progress is greater than the third preset threshold, then the user will be classified into a user group tagged as high intent. If it is determined that the duration of the page stay is less than or equal to the second preset threshold and / or the form completion progress is less than or equal to the third preset threshold, then the user will be classified into a user group labeled as having ordinary intentions.

3. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, Determine whether the preset dynamic adjustment conditions are met, including: Obtain the current system time and retrieve the preset adjustment period, wherein the adjustment period defines the time interval for dynamic adjustment; Calculate the time difference between the current time and the time when the dynamic adjustment was last successfully executed; Determine whether the time difference has reached or exceeded the adjustment period; If the time difference reaches or exceeds the adjustment period, it is determined that the dynamic adjustment conditions are met.

4. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, Calculate and update the user traffic splitting ratio based on the conversion rate data, including: Obtain the real-time conversion rate of each insurance application process version within the current statistics window, wherein the real-time conversion rate is the ratio of the number of users who successfully completed the insurance application process version to the total number of users accessing the insurance application process version; The weight of insurance products is determined based on the types of insurance products covered by each version of the insurance application process. Based on the real-time conversion rate and the insurance type weight of each insurance application process version, the weighted conversion rate corresponding to each insurance application process version is determined by the weighted conversion rate calculation formula. Based on the weighted conversion rate and the sum of the weighted conversion rates of all insurance application process versions, the user diversion ratio corresponding to each insurance application process version is determined by the diversion ratio formula.

5. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 4, characterized in that, The weight of insurance products is determined based on the types of insurance products covered by each version of the application process, including: Based on the insurance product type, the basic weight corresponding to the insurance product type is obtained according to the predefined insurance product weight configuration table; Obtain the current system time and determine whether the current time is within a preset target period, wherein the target period includes holidays, large-scale promotional events, and high-value time periods of the day; If it is determined that the current time is within the preset target period, the final effective insurance type weight is determined based on the current time and the preset time period-weight mapping table. The time period-weight mapping table stores the mapping relationship between a specific time period and the corresponding insurance type weight in the form of key-value pairs. If it is determined that the current time is not within the preset target period, then the basic weight will be used as the final effective insurance type weight.

6. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, Based on the user segmentation and the currently effective user diversion ratio, a corresponding insurance application process version is assigned to the user, including: Based on the segmentation tags corresponding to the user segments, determine one or more candidate insurance application process versions; Obtain the currently effective user traffic splitting ratio, wherein the user traffic splitting ratio defines the proportion of user traffic that should be allocated to each candidate version at the current moment; Based on the currently effective user diversion ratio, a consistent hashing algorithm is used to bucket user identifiers, thereby assigning a unique target insurance process version to the current user from the candidate insurance process versions.

7. The method for dynamic diversion of insurance application process versions based on user segmentation according to claim 1, characterized in that, Before allocating insurance process versions to subsequent users based on the updated user diversion ratio, the method further includes: After calculating a preliminary user diversion ratio adjustment plan based on the conversion rate data, one or more target insurance process versions with a planned increase in the diversion ratio are identified from the plan. For each of the target underwriting process versions proposed for improvement, a real-time compliance verification process will be executed, specifically including: Obtain the user ID of the user currently to be diverted, and query the corresponding user attribute data based on the user ID, wherein the user attribute data includes at least geographical location and age; Based on the user's geographical location, age, and the identifier of the target insurance application process version as input, the compliance verification service is invoked. By querying the pre-set compliance rule library, the verification is performed to determine whether the target insurance application process version meets all mandatory compliance requirements for the corresponding user's location, insurance type, and user attributes. If the verification passes, it is determined that the target insurance application process version meets the compliance requirements, and the operation to increase its diversion ratio is allowed. If the verification fails, it is determined that the target insurance application process version does not meet the compliance requirements, and the current increase in the diversion ratio for the target insurance application process version is cancelled.

8. A dynamic diversion system for the insurance application process based on user segmentation, characterized in that, include: The data acquisition module is used to respond to users' insurance application requests and acquire user attribute data and real-time behavior data; The user segmentation module is used to divide the user into predefined user segments based on the user attribute data and the real-time behavior data. The version allocation module is used to allocate the corresponding insurance process version to the user based on the user group and the current effective user diversion ratio; The data collection and statistics module is used to collect and statistically analyze the conversion rate data of each insurance application process version in real time. The strategy decision module is used to determine whether preset dynamic adjustment conditions are met; when it is determined that the dynamic adjustment conditions are met, the user diversion ratio is calculated and updated based on the conversion rate data. The version allocation module is also used to allocate insurance process versions to subsequent users based on the updated user diversion ratio.

9. An electronic device comprising a processor, a communication interface, a memory, and a communication bus, wherein, The processor, the communication interface, and the memory communicate with each other via the communication bus, characterized in that... The memory is used to store computer programs; The processor is configured to execute the dynamic triage method for the insurance application process based on user segmentation as described in any one of claims 1 to 7 by running the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to execute the dynamic diversion method for the insurance application process based on user segmentation as described in any one of claims 1 to 7 when it is run.

Images