CN-121998768-A - Intelligent insurance client dynamic region division method based on multi-source positioning compensation

CN121998768ACN 121998768 ACN121998768 ACN 121998768ACN-121998768-A

Abstract

The application discloses an intelligent insurance client dynamic region dividing method based on multi-source positioning compensation, which comprises the steps of S1, obtaining client multi-source positioning data according to a cooperation protocol with clients and preprocessing, S2, constructing client dynamic behavior portraits by adopting a cluster analysis method based on the preprocessed data in S1, S3, generating personalized dynamic geofences according to the client dynamic behavior portraits constructed in S2, S4, calculating client density according to the client dynamic behavior portraits constructed in S2 and the client dynamic geofences constructed in S3, and dividing map grids based on the client density, and S5, generating insurance service strategies based on the client dynamic behavior portraits constructed in S2 and the map grids divided in S4. The response speed of the trans-regional service is improved by an order of magnitude, the service period is greatly shortened, the operation cost is obviously reduced, and the whole customer operation efficiency is finally driven to take a brand new step.

Inventors

LIU TIANYI
YANG LIU
ZHENG MINGMING

Assignees

中国人寿保险股份有限公司新疆维吾尔自治区分公司

Dates

Publication Date: 20260508
Application Date: 20251211

Claims (8)

1. An intelligent insurance client dynamic region dividing method based on multi-source positioning compensation is characterized by comprising the following steps: s1, acquiring multi-source positioning data of a client according to a cooperation protocol with the client and preprocessing the multi-source positioning data; s2, constructing a customer dynamic behavior portrait by adopting a cluster analysis method based on the data preprocessed in the S1; s3, generating personalized dynamic geofences according to the client dynamic behavior portraits constructed in the S2; S4, calculating the client density according to the client dynamic behavior image constructed in the S2 and the client dynamic geofence constructed in the S3, and dividing the map grid based on the client density; S5, generating an insurance service policy based on the client dynamic behavior portraits constructed in the S2 and the map grids divided in the S4.
2. The intelligent partitioning method for insurance client dynamic areas based on multi-source positioning compensation according to claim 1, further comprising: And S6, when the client is at the grid juncture, performing cross-regional service collaboration and intelligent decision support.
3. The intelligent division method for dynamic regions of insurance clients based on multi-source positioning compensation according to claim 1, wherein the step of obtaining client multi-source positioning data and preprocessing in S1 comprises: S1.1, adopting GPS, beidou, wi-Fi signals and a base station positioning multi-source technology to perform collaborative data acquisition so as to ensure that the positioning capability can be maintained when a single signal source fails; S1.2, setting positioning data to collect according to preset frequency, and supporting a self-adaptive frequency adjustment strategy, namely, increasing the frequency when detecting that a client is in a high-speed moving state, and reducing the frequency when the client is stationary for a long time so as to optimize the energy consumption of a terminal; S1.3, denoising and fusing the acquired original coordinate data to generate a continuous, smooth and credible client track sequence.
4. The intelligent partitioning method for insurance client dynamic area based on multi-source positioning compensation according to claim 3, wherein the constructing client dynamic behavior image in S2 includes: S2.1, analyzing continuous historical track points of clients by using a density clustering algorithm, identifying an active hot spot area of the continuous historical track points, and setting differential parameters according to urban density types; S2.2, analyzing the stay time, the access frequency and the access time of the client in each hot spot area, and deducing the functional attribute of the area so as to construct the dynamic behavior portrait of the client.
5. The intelligent partitioning method for insurance-customer dynamic area based on multi-source positioning compensation according to claim 4, wherein the dynamic geofence in S3 is a personalized dynamic geofence generated based on each active hot spot area obtained in S2.1, wherein the edges of the geofence are matched with the actual building or area boundary, and the range and position of the fence are adjusted according to the change of customer behavior or seasonal law.
6. The intelligent partitioning method for dynamic area of insurance client based on multi-source positioning compensation according to claim 5, wherein the step of performing map meshing in S4 includes: s4.1, dividing the map into a grid system with multi-level resolution by adopting a space index algorithm such as a quadtree or an H3 grid; S4.2, dynamically adjusting the grid granularity according to the real-time client density in each grid, adopting fine-granularity grids in the central area of the city with dense clients to realize accurate division, and adopting coarse-granularity grids in suburbs with sparse clients to improve the calculation efficiency.
7. The intelligent partitioning method for dynamic area of insurance client based on multi-source positioning compensation according to claim 6, wherein said generating an insurance service policy in S5 includes: s5.1, distributing weight coefficients according to the functional attributes of the area inferred in the S2.2, and comprehensively considering the customer grade and the product type; and S5.2, calculating the attribution and service priority sequence of the service area according to the grid where the client is located in real time based on the weight coefficient distributed in the S5.1 and the grid divided in the S4.
8. The intelligent partitioning method for dynamic regions of insurance clients based on multi-source positioning compensation according to claim 2, wherein said S6 comprises: s6.1, when a client is positioned at a grid or an area juncture, a trans-regional service request is provided, a trans-regional service cooperative mechanism is started by the system, and service path planning and intelligent dispatch of service personnel are calculated based on real-time positions, service team loads and traffic road condition information; S6.2, setting a visual decision board to display the state of the cross-regional service work order, the resource allocation condition and the system recommended solution, supporting manual intervention and final decision, and ensuring that the key service request is responded most efficiently.

Description

Intelligent insurance client dynamic region division method based on multi-source positioning compensation Technical Field The application relates to the technical field of intersection of artificial intelligence and insurance pricing, in particular to an intelligent division method for dynamic areas of insurance clients based on multi-source positioning compensation. Background Currently, the division of the customer area in the insurance industry mainly depends on the traditional administrative division or static geographic division method, and the fixed area allocation is generally performed based on the home address or the work unit address registered by the customer. The division mode is convenient to manage, but has obvious limitations that on one hand, the actual activity range and mobility of a client are difficult to reflect, so that resource allocation and actual demand are misplaced, and on the other hand, static division easily causes service response lag or region overlapping due to frequent changes of the living or working place of the client, so that service efficiency and accurate marketing effect are affected. In the prior art practice, part of insurance companies try to introduce a GPS positioning technology to dynamically identify the client position so as to make up for the defects of the traditional dividing method. However, relying on single GPS positioning still faces various challenges, including limited positioning accuracy, high positioning deviation, limited signal coverage and stability, poor positioning effect in remote areas, underground spaces or severe weather conditions, and significant increase of energy consumption of terminal equipment due to continuous use of GPS, which affects user experience and equipment endurance. In addition, the situation of behavior scene analysis is lacking simply by relying on the position data, and it is difficult to comprehensively grasp the actual insurance needs and service opportunities of the clients. Disclosure of Invention The application aims to provide an intelligent insurance client dynamic region dividing method based on multi-source positioning compensation, which comprises the following specific technical scheme: A method for intelligently dividing an insurance client dynamic region based on multi-source positioning compensation comprises the steps of S1, obtaining multi-source positioning data of clients according to a cooperation protocol with the clients and preprocessing the multi-source positioning data, S2, constructing client dynamic behavior portraits by adopting a clustering analysis method based on the preprocessed data in S1, S3, generating personalized dynamic geofences according to the client dynamic behavior portraits constructed in S2, S4, calculating client density according to the client dynamic behavior portraits constructed in S2 and the client dynamic geofences in S3, dividing map grids based on the client density, and S5, generating insurance service strategies based on the client dynamic behavior portraits constructed in S2 and the map grids divided in S4. S6, when the client is at the grid juncture, cross-regional service collaboration and intelligent decision support are carried out. The method comprises the steps of S1.1 acquiring multi-source positioning data of a client and preprocessing the multi-source positioning data, wherein the GPS, the Beidou, wi-Fi signals and a base station positioning multi-source technology are adopted for collaborative data acquisition to ensure that positioning capability can still be maintained when a single signal source fails, S1.2 acquiring the positioning data according to preset frequency, supporting a self-adaptive frequency adjustment strategy, namely improving frequency when the client is detected to be in a high-speed moving state and reducing frequency when the client is stationary for a long time, so as to optimize terminal energy consumption, and S1.3 denoising and fusing the acquired original coordinate data to generate a continuous, smooth and reliable client track sequence. The step S2.2 is used for analyzing the stay time, the access frequency and the access time of the client in each hot spot area so as to deduce the functional attribute of the area and construct the dynamic behavior portrait of the client. The dynamic geofence in S3 is a personalized dynamic geofence generated based on each active hot spot area obtained in S2.1, wherein the edges of the geofence are matched with the boundaries of actual buildings or areas, and the range and the position of the fence are adjusted according to the change of customer behaviors or seasonal rules. The step S4 is that the map is divided into a grid system with multi-level resolution by adopting a space index algorithm such as a quadtree or an H3 grid and the like, and the step S4.2 is that the grid granularity is dynamically adjusted according to the real-time client density in each grid, fine-granularity grids a