Insurance customer dynamic area intelligent division method based on multi-source positioning compensation

Abstract

The invention discloses an insurance customer dynamic area intelligent division method based on multi-source positioning compensation. The method comprises the following steps: S1, acquiring customer multi-source positioning data according to a cooperation protocol with a customer, and preprocessing the customer multi-source positioning data; s2, on the basis of the data preprocessed in the S1, a clustering analysis method is adopted to construct a client dynamic behavior portrait; s3, generating a personalized dynamic geofence according to the dynamic behavior portrait of the customer constructed in the S2; s4, calculating customer density according to the customer dynamic behavior portrait constructed in the S2 and the customer dynamic geofence constructed in the S3, and performing map grid division based on the customer density; and S5, generating an insurance service strategy based on the customer dynamic behavior portrait constructed in the S2 and the map grid divided in the S4. Therefore, the response speed of the cross-regional service is improved in order of magnitude, the service period is greatly shortened, the operation cost is remarkably reduced, and finally the overall customer operation efficiency is driven to reach a brand new step.

Description

Technical Field

[0001] This application relates to the interdisciplinary field of artificial intelligence and insurance pricing, specifically to a method for intelligently dividing the dynamic regions of insurance customers based on multi-source positioning compensation. Background Technology

[0002] Currently, the insurance industry's customer regional division mainly relies on traditional administrative divisions or static geographical methods, typically using the customer's registered home address or workplace address as the basis for fixed area allocation. While this method is convenient for management, it has significant limitations: firstly, it fails to reflect the actual range and mobility of customers' activities, leading to a mismatch between resource allocation and actual needs; secondly, because customers' residences or workplaces may change frequently, static division can easily cause service response delays or overlapping areas, affecting service efficiency and the effectiveness of targeted marketing.

[0003] In existing technological practices, some insurance companies have attempted to introduce GPS positioning technology for dynamic customer location identification to compensate for the shortcomings of traditional segmentation methods. However, relying solely on GPS positioning still faces several challenges: First, positioning accuracy is limited, especially in densely built-up areas or indoor environments where signals are easily interfered with, resulting in significant positioning deviations; second, due to limitations in signal coverage and stability, positioning performance is poor in remote areas, underground spaces, or in adverse weather conditions; third, continuous use of GPS significantly increases the energy consumption of terminal devices, affecting user experience and device battery life. Furthermore, relying solely on location data without behavioral scenario analysis makes it difficult to fully grasp customers' actual insurance needs and service opportunities. Summary of the Invention

[0004] The purpose of this application is to provide a method for intelligent dynamic regional division of insurance customers based on multi-source positioning compensation. The specific technical solution is as follows:

[0005] A method for intelligent dynamic area segmentation of insurance customers based on multi-source positioning compensation includes: S1, obtaining multi-source positioning data of customers according to the cooperation agreement with the customers and performing preprocessing; S2, constructing a dynamic behavior profile of customers based on the preprocessed data in S1 using cluster analysis; S3, generating personalized dynamic geofences based on the dynamic behavior profiles of customers constructed in S2; S4, calculating customer density based on the dynamic behavior profiles of customers constructed in S2 and the dynamic geofences of customers in S3, and dividing the map into grids based on the customer density; S5, generating insurance service strategies based on the dynamic behavior profiles of customers constructed in S2 and the map grids divided in S4.

[0006] It also includes: S6, providing cross-regional service collaboration and intelligent decision support when customers are at grid boundaries.

[0007] When acquiring and preprocessing multi-source customer positioning data in S1, the process includes: S1.1, using GPS, BeiDou, Wi-Fi signals, and base station positioning technologies for collaborative data acquisition to ensure positioning capability is maintained even when a single signal source fails; S1.2, setting the positioning data to be collected at a preset frequency, while supporting an adaptive frequency adjustment strategy, i.e., increasing the frequency when the customer is detected to be moving at high speed and decreasing the frequency when the customer is stationary for a long time, thereby optimizing terminal power consumption; S1.3, performing noise reduction and fusion processing on the collected raw coordinate data to generate a continuous, smooth, and reliable customer trajectory sequence.

[0008] When constructing a dynamic customer behavior profile in S2, the following steps are taken: S2.1, using density clustering algorithm to analyze the continuous historical trajectory points of customers, identify their activity hotspots, and set differentiated parameters according to the city density type; S2.2, analyzing the length of stay, frequency of visits, and time of visit of customers in each hotspot area to infer the functional attributes of the area, thereby constructing a dynamic customer behavior profile.

[0009] The dynamic geofence in S3 is a personalized dynamic geofence generated based on each activity hotspot area obtained in S2.1; the edges of the geofence match the actual building or area boundaries, and the range and location of the geofence are adjusted according to changes in customer behavior or seasonal patterns.

[0010] The map grid division process in S4 includes: S4.1, using spatial indexing algorithms such as quadtrees or H3 grids to divide the map into a multi-level resolution grid system; S4.2, dynamically adjusting the grid granularity according to the real-time customer density within each grid, using fine-grained grids in densely populated urban centers to achieve accurate division, and using coarse-grained grids in sparsely populated suburban areas to improve computational efficiency.

[0011] When generating insurance service strategies in S5, the following steps are included: S5.1, assigning weight coefficients based on the functional attributes of the area inferred in S2.2, and taking into account customer level and product type; S5.2, calculating the service area affiliation and service priority sequence based on the weight coefficients assigned in S5.1 and the grids divided in S4, according to the grid where the customer is located in real time.

[0012] S6 includes: S6.1 When a customer is located at the boundary of a grid or region and submits a cross-regional service request, the system initiates a cross-regional service collaboration mechanism, calculates service route planning and intelligently dispatches service personnel based on real-time location, service team load and traffic conditions; S6.2 Sets up a visual decision dashboard to display the status of cross-regional service work orders, resource allocation and system-recommended solutions, supports manual intervention and final decision-making, and ensures that critical service requests receive the most efficient response.

[0013] The beneficial effects of this application lie in its integration of multi-source information such as GPS, Wi-Fi signals, base station data, and IoT sensors, and the use of advanced data fusion algorithms (such as Kalman filtering) for cross-validation and compensation correction. This fundamentally overcomes the inherent defects of single GPS technology, such as signal attenuation and positioning drift, in complex scenarios like indoors, underground, or urban canyons. This not only solves the problem of lag and distortion in traditional static address information but also achieves high-precision, high-reliability identification and scenario-based analysis of customers' multi-dimensional permanent locations, including their "workplace," "residence," and "high-frequency activity areas," providing a solid data foundation for precise services. It completely breaks away from the static division model relying on fixed administrative boundaries. By introducing an artificial intelligence model, the system can continuously learn and analyze customer location trajectory data, automatically identifying their activity patterns and regularities. This makes regional division no longer pre-set and unchanging, but an intelligent process that can perceive customer location movement in real time and dynamically adjust service area assignments. The system possesses high adaptability and flexibility, capable of responding instantly to various location changes such as customer migration, commuting, and temporary business trips, ensuring that the division of service responsibility areas always remains highly consistent with the actual situation of the customer. Based on the dynamic and precise segmentation results, the system can achieve on-demand, precise scheduling and optimal allocation of service resources (such as claims adjusters and sales representatives). This effectively solves the structural contradiction of uneven task load and resource idleness and shortage in different regions under the traditional model, thus achieving a qualitative leap in the overall efficiency of service resource utilization. At the same time, since customers can always be efficiently associated with the most suitable service unit and receive fast and accurate service responses, their service experience and satisfaction are also significantly improved in tandem, forming a virtuous cycle of mutual promotion between resource optimization and experience improvement. Traditional models often suffer from response delays when handling cross-regional business due to unclear responsibilities and complex processes. This invention completely breaks down information barriers between regions by constructing a unified dynamic map and intelligent task routing mechanism. The system can intelligently identify cross-regional service needs and automatically select the optimal service node or initiate collaborative processing, thereby eliminating the pain point of service delays at the mechanism level. This results in an order-of-magnitude improvement in the response speed of cross-regional services, greatly shortens the service cycle, significantly reduces operating costs, and ultimately drives overall customer management efficiency to a new level. Attached Figure Description

[0014] Figure 1 This is a flowchart illustrating the application process. Detailed Implementation

[0015] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to specific embodiments and accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of this application. Furthermore, descriptions of well-known structures and technologies are omitted in the following description to avoid unnecessarily obscuring the concepts of this application.

[0016] like Figure 1 As shown, a method for intelligent dynamic regional division of insurance customers based on multi-source positioning compensation includes:

[0017] S1. Obtain and preprocess customer multi-source positioning data according to the cooperation agreement with the customer. Specifically, obtaining and preprocessing customer multi-source positioning data includes: S1.1. Using GPS, Beidou, Wi-Fi signals, and base station positioning technologies for collaborative data acquisition to ensure positioning capability is maintained even when a single signal source fails (e.g., GPS signal loss due to entering indoors); S1.2. Setting the positioning data to be collected at a preset frequency, while supporting an adaptive frequency adjustment strategy, i.e., increasing the frequency when the customer is detected to be moving at high speed and decreasing the frequency when the customer is stationary for a long time, thereby optimizing terminal power consumption; S1.3. Denoising (e.g., removing drift points that are significantly off the path) and fusing the collected raw coordinate data, and generating a continuous, smooth, and more reliable customer trajectory sequence through algorithms (e.g., Kalman filtering).

[0018] S2. Based on the preprocessed data in S1, cluster analysis is used to construct a dynamic customer behavior profile. Specifically, constructing a dynamic customer behavior profile includes: S2.1. Analyzing continuous historical trajectory points of customers using density clustering algorithms to identify their activity hotspots, and setting differentiated parameters based on city density type (e.g., city center vs. suburbs) (e.g., min_samples=5, eps=100 meters); S2.2. Analyzing the customer's dwell time, visit frequency, and visit time (e.g., weekday daytime / nighttime / weekend) in each hotspot area to infer the functional attributes of the area (e.g., "primary workplace," "core residential area," "secondary leisure area"), thereby constructing a dynamic customer behavior profile.

[0019] S3. Generate personalized dynamic geofences based on the customer dynamic behavior profiles constructed in S2. Specifically, the dynamic geofences are personalized dynamic geofences generated based on each activity hotspot area obtained in S2.1; the edges of the geofences match the actual building or area boundaries, and the range and location of the geofences are adjusted according to changes in customer behavior (such as changes in work location) or seasonal patterns (such as the emergence of new high-frequency points on weekends).

[0020] S4. Calculate customer density based on the customer dynamic behavior profile constructed in S2 and the customer dynamic geofencing in S3, and then perform map grid division based on the customer density. Specifically, map grid division includes: S4.1, using spatial indexing algorithms such as quadtrees or H3 grids to divide the map into a multi-level resolution grid system; S4.2, dynamically adjusting the grid granularity according to the real-time customer density within each grid, using fine-grained grids in densely populated urban centers for accurate division, and coarse-grained grids in sparsely populated suburban areas to improve computational efficiency.

[0021] S5. Generate insurance service strategies based on the dynamic customer behavior profile built in S2 and the map grid divided in S4. Specifically, generating insurance service strategies includes: S5.1. Assigning weight coefficients based on the functional attributes of the area inferred in S2.2 (e.g., primary residence has higher weight, followed by workplace), and comprehensively considering customer level and product type; S5.2. Based on the weight coefficients assigned in S5.1 and the grid divided in S4, calculating the service area affiliation and service priority sequence according to the customer's real-time grid location. For example, prioritizing assigning customers to the service team corresponding to their "primary activity area," and generating refined strategies such as "prioritizing contacting Team A during weekdays and Team B at night and on weekends."

[0022] S6. When a customer is located at a grid boundary, cross-regional service collaboration and intelligent decision support are implemented. Specifically, this includes: S6.1. When a customer is located at a grid or regional boundary and submits a cross-regional service request, the system initiates a cross-regional service collaboration mechanism. Based on real-time location, service team load, and traffic conditions, it calculates service route planning and intelligently dispatches service personnel. S6.2. A visual decision dashboard is set up to display the status of cross-regional service work orders, resource allocation, and system-recommended solutions, supporting manual intervention and final decision-making to ensure that critical service requests receive the most efficient response.

Claims

1. A method for intelligent dynamic regional division of insurance customers based on multi-source positioning compensation, characterized in that, include: S1. Obtain multi-source positioning data from customers according to the cooperation agreement with the customers and perform preprocessing; S2. Based on the preprocessed data in S1, a cluster analysis method is used to construct a dynamic customer behavior profile; S3. Generate personalized dynamic geofences based on the customer dynamic behavior profile constructed in S2; S4. Calculate customer density based on the customer dynamic behavior profile constructed in S2 and the customer dynamic geofence in S3, and divide the map grid based on the customer density. S5. Generate an insurance service strategy based on the customer dynamic behavior profile constructed in S2 and the map grid divided in S4.

2. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 1, characterized in that, Also includes: S6. When customers are located at grid boundaries, cross-regional service collaboration and intelligent decision support are provided.

3. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 1, characterized in that, The process of acquiring and preprocessing customer multi-source location data in S1 includes: S1.

1. Multi-source technologies such as GPS, Beidou, Wi-Fi signals and base station positioning are used for coordinated data collection to ensure that positioning capability can still be maintained when a single signal source fails. S1.2 Set the location data to be collected at a preset frequency, and support an adaptive frequency adjustment strategy, that is, increase the frequency when the customer is detected to be moving at high speed and decrease the frequency when the customer is stationary for a long time, so as to optimize the terminal power consumption. S1.

3. The collected raw coordinate data is denoised and fused to generate a continuous, smooth and reliable customer trajectory sequence.

4. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 3, characterized in that, The process of constructing a dynamic customer behavior profile in S2 includes: S2.1 Analyze the continuous historical trajectory points of customers using density clustering algorithm, identify their activity hotspot areas, and set differentiated parameters according to the city density type; S2.2 Analyze the length of time customers stay, frequency of visits, and time of arrival in each hotspot area to infer the functional attributes of the area and thus construct a dynamic behavioral profile of the customer.

5. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 4, characterized in that, The dynamic geofence in S3 is a personalized dynamic geofence generated based on each activity hotspot area obtained in S2.

1. The edges of the geofence match the actual building or area boundaries, and the range and location of the geofence are adjusted according to changes in customer behavior or seasonal patterns.

6. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 5, characterized in that, The map grid division process in S4 includes: S4.

1. Use spatial indexing algorithms such as quadtrees or H3 grids to divide the map into a multi-level resolution grid system; S4.

2. Dynamically adjust the grid granularity based on the real-time customer density within each grid. Use fine-grained grids in densely populated urban centers to achieve accurate partitioning, and use coarse-grained grids in sparsely populated suburban areas to improve computational efficiency.

7. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 6, characterized in that, The generation of the insurance service strategy in S5 includes: S5.1 Assign weight coefficients to the functional attributes of the area inferred in S2.2, taking into account customer level and product type; S5.

2. Based on the weight coefficients allocated in S5.1 and the grids divided in S4, calculate the service area affiliation and service priority sequence according to the grid where the customer is located in real time.

8. The method for intelligent division of dynamic regions for insurance customers based on multi-source positioning compensation as described in claim 2, characterized in that, S6 includes: S6.1 When a customer is located at the boundary of a grid or region and submits a cross-regional service request, the system will activate the cross-regional service collaboration mechanism, calculate service route planning and intelligently dispatch service personnel based on real-time location, service team load and traffic information. S6.2 Set up a visual decision dashboard to display the status of cross-regional service work orders, resource allocation, and system-recommended solutions, supporting human intervention and final decision-making to ensure that critical service requests receive the most efficient response.

Images