CN-122023020-A - Method, device, storage medium and program product for processing data of learning young insurance claim

CN122023020ACN 122023020 ACN122023020 ACN 122023020ACN-122023020-A

Abstract

The embodiment of the application provides a method, equipment, a storage medium and a program product for processing data of learning young insurance claims. On the basis of acquiring the claim data, the pre-trained rule verification model carries out structural verification on the claim data based on preset business rules (such as dangerous compliance and cost rationality) to generate a verification result, and finally, a claim processing instruction is generated according to the verification result so as to realize automatic processing of young risk cases. The technical means obviously reduces case processing time and labor cost by replacing a manual auditing process, reduces false claim settlement risk by multi-dimensional checking logic (such as medical bill compliance checking and personal injury approval rules), and improves insurance claim settlement efficiency and accuracy.

Inventors

Shi Anjing
LU CHANGSONG
LI ZHENGRONG

Assignees

中国人民财产保险股份有限公司

Dates

Publication Date: 20260512
Application Date: 20251229

Claims (10)

1. A method of processing data for a child learning claim, the method comprising: Acquiring claim data, wherein the claim data comprises case data and medical bill images of young learning insurance cases submitted by users; Invoking a pre-trained rule verification model to carry out compliance verification on the claim data, and generating a verification result; And generating an claim settlement instruction according to the verification result, wherein the claim settlement instruction is used for indicating a claim settlement flow for settling the young risk cases.
2. The method of claim 1, wherein invoking the pre-trained rule check model to compliance check the claim data generates a check result comprises: invoking the pre-trained rule verification model to carry out admission rule verification on the case data, and generating an admission verification result; invoking the pre-trained rule verification model to perform human injury verification rules on the medical bill image to generate a human injury verification result; And verifying the human injury verification result according to the admission verification result to generate the verification result.
3. The method of claim 2, wherein the pre-trained rule checking model comprises a first image recognition module and a rule checking module; The invoking of the pre-trained rule verification model performs human injury verification of the medical bill image to generate a human injury verification result, and the method comprises the following steps: Invoking the first image recognition module to extract text field information in the medical bill image and generating medical bill structured data; And calling the rule verification module to perform human injury verification rule verification on the medical bill structured data, and generating the human injury verification result.
4. The method of claim 3, further comprising, after said invoking said rule checking module to perform a human injury verification rule check on said medical ticket structured data to generate said human injury verification result: displaying the human injury verification result on an interactive display unit; Acquiring real-time feedback data, wherein the real-time feedback data is input by insurance claims personnel through the interactive display unit based on the human injury verification result; and adjusting parameters of the human injury verification rule according to the historic young risk case processing result and the real-time feedback data.
5. The method of claim 1, further comprising, prior to the invoking the pre-trained rule check model to compliance check the claim data to generate a check result: Calling a multi-mode data fusion model to verify the medical bill image to generate an image verification result; If the image verification result is passed, executing the step of calling a pre-trained rule verification model to carry out compliance verification on the claim data and generating a verification result; And if the image verification result is that the image verification result is not passed, sending medical bill image verification non-passing information to the terminal equipment of the user.
6. The method of claim 5, wherein the multimodal data fusion model includes a second image recognition module and a text recognition module; the calling the multi-mode data fusion model to verify the medical bill image to generate an image verification result comprises the following steps: invoking the second image recognition module to perform feature extraction on the medical bill image to generate image features of the medical bill image, structural data of medical treatment records and diagnosis proving data of the medical treatment records; invoking the text recognition module to extract text field information in the medical bill image and generating text characteristics of the medical bill image; Generating a bill source verification result according to the matching degree of the image characteristics of the medical bill image and the text characteristics of the medical bill image; generating a medical rationality verification result according to the matching degree of the structured data of the medical treatment record and the diagnosis proving data of the medical treatment record; And summarizing the bill source verification result and the medical rationality verification result to obtain the image verification result.
7. A data processing apparatus for learning young insurance claims, comprising: the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring claim data, wherein the claim data comprises case data of young learning danger cases submitted by users and medical bill images; The processing module is used for calling a pre-trained rule verification model to carry out compliance verification on the claim data and generating a verification result; and the generation module is used for generating an claim settlement instruction according to the verification result, wherein the claim settlement instruction is used for indicating a claim settlement flow for settling the young risk cases.
8. An electronic device comprising a processor and a memory communicatively coupled to the processor; The memory stores computer-executable instructions; The processor executes computer-executable instructions stored in the memory to implement the method of any one of claims 1 to 6.
9. A computer readable storage medium having stored therein computer executable instructions which when executed by a processor are adapted to carry out the method of any one of claims 1 to 6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the method of any one of claims 1 to 6.

Description

Method, device, storage medium and program product for processing data of learning young insurance claim Technical Field The application relates to the technical field of artificial intelligence, in particular to a method, equipment, a storage medium and a program product for processing data of learning young insurance claims. Background In the insurance industry, young-learning insurance (insurance against accident injury of minors) cases form a core part of accident insurance claim service due to the characteristics of high frequency and low amount. Due to the special features of minor claim settlement, the traditional claim settlement process needs to involve complex links such as guardian identity verification, medical bill authenticity verification, clinic records and risk reasons association analysis. In the prior art, the young risk learning and claim settlement process is mainly realized by combining OCR recognition with manual operation, and comprises manual material initial examination, medical bill examination, medical rationality examination, claim settlement and claim checking. However, the prior art scheme needs to rely on manual interaction for many times, and has the problem of low insurance claim settlement efficiency. Disclosure of Invention The data processing method, the device, the storage medium and the program product for the learning and young insurance claim settlement are used for solving the problem of low insurance claim settlement efficiency. In a first aspect, an embodiment of the present application provides a method for processing data of learning and adventure claim, where the data of claim includes case data of learning and adventure cases submitted by a user and medical bill images, a pre-trained rule verification model is called to perform compliance verification on the data of claim to generate a verification result, and a claim processing instruction is generated according to the verification result and is used for indicating a claim settlement flow for performing claim settlement on the learning and adventure cases. In a possible implementation manner, the calling the pre-trained rule verification model to carry out compliance verification on the claim data to generate a verification result comprises the steps of calling the pre-trained rule verification model to carry out admission rule verification on the case data to generate an admission verification result, calling the pre-trained rule verification model to carry out personnel injury verification rule verification on the medical bill image to generate a personnel injury verification result, and collecting the admission verification result and the personnel injury verification result to obtain the verification result. In a possible implementation manner, the pre-trained rule verification model comprises a first image identification module and a rule verification module, wherein the invoking of the pre-trained rule verification model carries out human injury verification rule verification on the medical bill image to generate a human injury verification result, the invoking of the first image identification module extracts text field information in the medical bill image to generate medical bill structured data, and the invoking of the rule verification module carries out human injury verification rule verification on the medical bill structured data to generate the human injury verification result. In a possible implementation manner, after the rule verification module is called to verify the human injury verification rule on the medical bill structured data and generate the human injury verification result, the method further comprises the steps of displaying the human injury verification result on an interactive display unit, acquiring real-time feedback data, wherein the real-time feedback data are input by insurance claims personnel through the interactive display unit based on the human injury verification result, and adjusting parameters of the human injury verification rule according to historic young risk case processing results and the real-time feedback data. In a possible implementation manner, before the invoking of the pre-trained rule verification model to carry out compliance verification on the claim data and generate a verification result, the method further comprises the steps of invoking a multi-mode data fusion model to verify the medical bill image and generate an image verification result, executing the invoking of the pre-trained rule verification model to carry out compliance verification on the claim data and generate a verification result if the image verification result is passed, and sending medical bill image verification non-passing information to terminal equipment of the user if the image verification result is not passed. In a possible implementation mode, the multi-mode data fusion model comprises a second image recognition module and a text recognition module, wherein the calling