Data evaluation method and device, computer equipment and storage medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to a data evaluation method and device, computer equipment and a storage medium, and the method comprises the steps: receiving an insurance processing request triggered by a user; judging whether the user is an insurance renewal user based on basic information extracted from the insurance processing request; if yes, querying related data of the user during the historical underwriting period; constructing portrait data based on the basic information and the related data; formatting the portrait data to obtain target data; performing information gain calculation processing on the target data based on the decision model, and constructing a target decision tree based on the obtained target information gain; performing risk assessment on the target data based on the target decision tree to obtain a risk assessment result; and generating an underwriting decision result based on the risk assessment result, and sending the underwriting decision result to the user. The method can be applied to insurance processing scenes in the financial science and technology field and the medical health insurance field, and the underwriting efficiency and accuracy of underwriting processing are improved based on the use of the decision model.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, and can be applied to the fields of financial technology, medical and health insurance, etc., and particularly relates to a data evaluation method and device, a computer device and a storage medium. BACKGROUND

[0002] In a traditional insurance underwriting service mode, insurance underwriting mainly relies on manual cooperation with fixed rules for underwriting processing, resulting in low underwriting efficiency and poor user experience. Specifically, the traditional underwriting method usually examines the insurance application based on pre-set simple rules (such as age range, premium limit, specific disease refusal list, etc.), and lacks the ability of deep analysis and dynamic adjustment of customer multi-dimensional information. This extensive underwriting method is difficult to accurately assess the risk level of customers, resulting in insufficient matching degree of underwriting results and actual risk conditions, which may not only lose high-quality business due to excessive underwriting (such as mis-rejecting low-risk customers), but also increase the risk of claims due to underwriting not being strict (such as mis-insuring high-risk customers), ultimately damaging the operating efficiency and market competitiveness of the insurance company.

[0003] For example, in the car insurance underwriting scene of the financial insurance field, the traditional method may only set a fixed rate discount according to the vehicle model, service life and customer historical claim frequency, without comprehensively analyzing key factors such as customer driving behavior data (such as annual average driving mileage, frequency of emergency braking), regional risk characteristics (such as accident-prone road sections). If the customer is a low-risk driver but the vehicle has been used for a long time, the traditional method may not be able to give reasonable rate discounts due to single rule restrictions, resulting in customer loss.

[0004] In the critical illness insurance underwriting scene of the medical and health insurance field, the traditional method may only make underwriting decisions based on the health questionnaire and medical history records filled out by the customer, without combining real-time health information such as wearable device data (such as heart rate, sleep quality), dynamic changes in physical examination reports. If the customer has potential health risks but has not yet reached the threshold of the traditional rules for rejection, the traditional method may mis-insure and cause subsequent claims disputes, which not only damages the company's interests, but also reduces the customer's trust in insurance services.

[0005] Therefore, it is urgent to provide an intelligent underwriting method to improve the underwriting efficiency and accuracy, realize the precise matching of risks and premiums, and optimize the overall performance of insurance services. SUMMARY

[0006] The purpose of the embodiments of the present application is to provide a data evaluation method and device, a computer device and a storage medium, to solve the technical problem of low underwriting efficiency and accuracy of existing insurance underwriting mainly relying on manual cooperation with fixed rules for underwriting processing.

[0007] In a first aspect, a data evaluation method is provided, comprising:

[0008] receiving a user triggered insurance application processing request; wherein the insurance application processing request carries basic information of the user;

[0009] extracting the basic information from the insurance application processing request, and judging whether the user is a renewal user based on the basic information;

[0010] if yes, querying related data of the user during a historical underwriting period from a preset database;

[0011] constructing portrait data of the user based on the basic information and the related data;

[0012] formatting the portrait data to obtain corresponding target data;

[0013] performing information gain calculation processing on the target data based on a preset decision model, and constructing a corresponding target decision tree based on the obtained target information gain;

[0014] performing risk assessment on the target data based on the target decision tree to obtain a corresponding risk assessment result;

[0015] generating a corresponding underwriting decision result based on the risk assessment result, and sending the underwriting decision result to the user.

[0016] In a second aspect, a data evaluation device is provided, comprising:

[0017] a receiving module configured to receive a user triggered insurance application processing request; wherein the insurance application processing request carries basic information of the user;

[0018] a first judging module configured to extract the basic information from the insurance application processing request, and judge whether the user is a renewal user based on the basic information;

[0019] a querying module configured to, if yes, query related data of the user during a historical underwriting period from a preset database;

[0020] a constructing module configured to construct portrait data of the user based on the basic information and the related data;

[0021] a first processing module configured to format the portrait data to obtain corresponding target data;

[0022] a second processing module configured to perform information gain calculation processing on the target data based on a preset decision model, and construct a corresponding target decision tree based on the obtained target information gain;

[0023] a first evaluation module configured to perform risk evaluation on the target data based on the target decision tree, to obtain a corresponding risk evaluation result;

[0024] a third processing module configured to generate a corresponding underwriting decision result based on the risk evaluation result, and send the underwriting decision result to the user.

[0025] In a third aspect, a computer device is provided, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the above data evaluation method when executing the computer program.

[0026] In a fourth aspect, a computer readable storage medium is provided, which stores a computer program, and the computer program implements the steps of the above data evaluation method when executed by a processor.

[0027] In the above data evaluation method, device, computer device, and storage medium, first, a user triggered underwriting processing request is received, wherein the underwriting processing request carries basic information of the user; then the basic information is extracted from the underwriting processing request, and it is judged whether the user is a renewal user based on the basic information; if yes, relevant data of the user during the historical underwriting period is queried from a preset database; then portrait data of the user is constructed based on the basic information and the relevant data; the portrait data is formatted to obtain corresponding target data; subsequently, the target data is subjected to information gain calculation processing based on a preset decision model, and a target decision tree is constructed based on the obtained target information gain; the target data is subjected to risk evaluation based on the target decision tree, to obtain a corresponding risk evaluation result; finally, a corresponding underwriting decision result is generated based on the risk evaluation result, and the underwriting decision result is sent to the user. Based on the above intelligent underwriting process, the application applies a decision model to automatically underwrite the underwriting processing request triggered by the user, significantly reduces the time and cost of manual review, reduces human intervention, and effectively improves the underwriting efficiency. Moreover, the application of the decision model makes the underwriting result more objective and transparent, and each step of decision making is supported by data, thereby improving the accuracy and reliability of the underwriting result. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the schemes in the present application, the drawings needed in the embodiments of the present application will be briefly introduced as follows. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0029] Figure 1 is an exemplary system architecture diagram in which the present application can be applied;

[0030] Figure 2 is a flow chart of one embodiment of a data evaluation method according to the present application;

[0031] Figure 3 is a structural schematic diagram of one embodiment of a data evaluation apparatus according to the present application;

[0032] Figure 4 is a structural schematic diagram of one embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application; the description and claims herein and the above description of drawings herein utilize terms such as "including" and "having" and variations thereof that are intended to be broad and encompass the terms "consisting of" and "consisting essentially of." The terms "first," "second," and the like, as used herein do not denote any order, quantity, or importance, but rather are used to distinguish one element from another.

[0034] Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrase "in an embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are they necessarily all directed to the same embodiment, or to a single alternative embodiment.

[0035] In order to make the technical personnel in the art better understand the scheme of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings.

[0036] As shown in Figure 1 , the system architecture 100 can include a terminal device 101, a network 102 and a server 103, the terminal device 101 can be a notebook computer 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is a medium for providing a communication link between the terminal device 101 and the server 103. The network 102 can include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0037] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0038] The terminal device 101 can be various electronic devices with display screens and supporting web browsing, in addition to the notebook computer 1011, the tablet computer 1012 or the mobile phone 1013, the terminal device 101 can also be an electronic book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer and a desktop computer, etc.

[0039] The server 103 can be a server providing various services, such as a background server providing support for the page displayed on the terminal device 101.

[0040] It should be noted that the data evaluation method provided by the embodiments of the present application is generally executed by the server / terminal device, and accordingly, the data evaluation apparatus is generally provided in the server / terminal device.

[0041] It should be understood that Figure 1 The number of terminal devices, networks and servers in

[0042] With reference to Figure 2 , a flowchart of one embodiment of the data evaluation method according to the present application is shown. The order of the steps in the flowchart can be changed according to different needs, and some steps can be omitted. The data evaluation method provided by the embodiments of the present application can be applied to any data evaluation scene related to the insurance processing, and then the data evaluation method can be applied to the products in these scenes, such as the insurance processing scenes in the financial technology field and the medical health insurance field. The data evaluation method comprises the following steps:

[0043] Step S201, receiving an insurance processing request triggered by a user; wherein the insurance processing request carries basic information of the user.

[0044] In the present embodiment, the electronic device (for example Figure 1The server / terminal device shown) can obtain a user-triggered insurance application processing request through a wired connection or a wireless connection. It should be noted that the wireless connection can include but is not limited to 3G / 4G / 5G connection, Wi-Fi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other now known or future developed wireless connection. The subject of the present application can be an intelligent underwriting system, also known as a data evaluation system, which can be referred to as a system. A user-friendly insurance page will be designed in the system, and the user can trigger the insurance processing request through the insurance page. The insurance page contains a series of input fields, which require the user to fill in basic information. These fields usually include but are not limited to: personal identity information: name, gender, age, ID number (as a unique identifier), contact information (phone number, email address), home address, etc. Insurance-related information: type of insurance, insurance amount, insurance period, etc. Among them, there will be clear prompts and instructions on the page to guide the user to fill in the information step by step. For example, through the mandatory item mark (such as an asterisk *), prompt the user which information is required, and provide examples of input formats (such as the format of the ID number). When the user fills in the information, the system will perform preliminary format verification, such as checking whether the ID number conforms to the national standard format, whether the phone number is a valid combination of digits, etc., to ensure that the basic information entered is correct in format.

[0045] Among them, the present application can be applied to the insurance processing scenarios in the fields of financial technology and medical health insurance. For example, in the personal property comprehensive insurance application scenario in the field of financial technology, the user-triggered insurance application processing request can include: the user wants to apply for property comprehensive insurance for a house in his name to prevent risks such as fire, theft, and natural disasters. The user's input basic information can include: personal identity information: name, ID number, contact information, home address, etc. Property information: property address, building area, building structure (such as reinforced concrete), property purpose (self-use / rental), property valuation, etc. Insurance needs include: insurance type: property comprehensive insurance coverage: fire, theft, flood, earthquake (optional) insurance amount: determined according to the property valuation, for example, 5 million yuan. Insurance period: 1 year. Other needs: whether to need additional liability (such as temporary rental expense compensation). The insurance application processing request includes: the user submits an insurance application in the system, requests the insurance company to provide property comprehensive insurance for his property, and hopes to obtain the premium quote and insurance details.

[0046] For example, in the personal medical insurance application scenario in the medical health insurance field, the user demand of the user triggered application processing request can include: the user hopes to apply for a medical insurance for himself / herself to cover the daily medical expenses, hospitalization expenses and major disease treatment expenses. The basic information input by the user can include: personal identity information: name, ID number, contact information, home address, etc. Health information: age, gender, height, weight, past medical history (such as hypertension, diabetes, etc.), family medical history, etc. Application demand includes: application type: medical insurance. Protection scope: outpatient expenses, hospitalization expenses, surgery expenses, drug expenses, major disease protection, etc. Insurance amount: for example, annual limit of 1 million yuan, major disease limit of 50 million yuan. Insurance period: 1 year (can be renewed). Other requirements: whether to need additional services (such as overseas medical treatment, second medical opinion, etc.). The application processing request includes: the user submits an application for insurance in the system, requests the insurance company to provide medical insurance protection, and hopes to obtain the premium quotation, insurance terms and health notification requirements.

[0047] In step S202, the basic information is extracted from the application processing request, and whether the user is a renewal user is determined based on the basic information.

[0048] In the embodiment, the basic information submitted by the user can be obtained by extracting information from the above-mentioned application processing request. Further, the process of determining whether the user is a renewal includes that the database of the insurance company stores the historical application records of all customers, which include but are not limited to the unique identification information (such as ID number) of the customer, the application time, the insurance expiration time, the insurance product type, the historical claim record, etc. When the user submits the basic information, the system extracts the unique identification information (such as ID number) provided by the user and queries in the historical records. The logic of the query is to check whether there is an application record related to the customer after the current insurance expiration time. The renewal determination includes that if the record of the user applying for insurance again after the current insurance expiration time is found, the system determines that the customer is a renewal user. This means that the customer has purchased insurance before, and now applies for insurance again after the insurance expires. The new customer determination includes that if no relevant record is found, the system determines that the customer is a new customer, i.e. a customer who applies for insurance in the insurance company for the first time. Further, the system records the determination result of the renewal or new customer as an important basis for subsequent data collection and processing process. For example, the renewal user can need to provide more historical data, and the new customer needs to collect all necessary information from the beginning.

[0049] In addition, if it is detected that the user is not a renewal user, i.e., a new customer, then the basic information input by the new user, and the on-site investigation data and the latest bank credit information of the new customer obtained by the salesperson are taken as inputs of the decision model, so as to complete the risk assessment and underwriting decision processing of the new user through the decision model.

[0050] In step S203, if yes, the relevant data of the user during the historical underwriting period are queried from a preset database.

[0051] In the embodiment, the database can refer to the database of the insurance company. The historical underwriting records of the customer can be retrieved in the database according to the unique identification information (such as the ID card number) provided by the user. These records include but are not limited to: loss records: the number of losses in the past period of time, the type of accidents (such as collision, natural disasters, etc.), the claim amount of each accident, etc. Target information: for customers of car insurance, including the use environment of the vehicle (such as the proportion of urban / rural driving, whether it is often driven in bad weather), the use frequency (such as the daily driving mileage, the number of use days per week), the use time (such as the age of the vehicle), etc. User basic information update: check whether the age, occupation, contact method, etc. of the user have changed to reflect the latest personal situation. Bank credit: through the interface with the bank system or the data query of the third-party credit agency, the latest credit score or credit report of the customer is obtained.

[0052] In step S204, the portrait data of the user is constructed based on the basic information and the relevant data.

[0053] In the embodiment, the specific implementation process of constructing the portrait data of the user based on the basic information and the relevant data will be further described in detail in subsequent specific embodiments, and will not be described too much here.

[0054] In step S205, the portrait data is formatted to obtain corresponding target data.

[0055] In the embodiment, the specific implementation process of formatting the portrait data to obtain the corresponding target data will be further described in detail in subsequent specific embodiments, and will not be described too much here.

[0056] In step S206, the target data is subjected to information gain calculation processing based on a preset decision model, and a corresponding target decision tree is constructed based on the obtained target information gain.

[0057] In this embodiment, the above-mentioned decision model can specifically adopt a decision tree model. The process of calculating information gain based on the decision model includes: Entropy calculation: The decision model first calculates the entropy of the entire data set. Entropy is an indicator to measure the uncertainty of the data set. The higher the entropy value, the greater the uncertainty of the data set. For example, for a two-classification problem (insurance or non-insurance), the entropy calculation formula is: Entropy = -p(insurance)*log2(p(insurance))-p(non-insurance)*log2(p(non-insurance)), where p(insurance) and p(non-insurance) are the proportions of insured and non-insured samples in the data set, respectively. Information gain calculation: For each feature (such as risk rate, credit score, occupation, etc.), calculate its information gain. Information gain indicates the degree of reduction in the uncertainty of the data set after using the feature for division. The information gain calculation formula is: Information gain = original entropy - weighted entropy after division. For example, for the risk rate feature, the system will divide the data set into multiple subsets according to different risk rate ranges, then calculate the entropy of each subset, and perform weighted summation to obtain the entropy after division.

[0058] The process of constructing the target decision tree includes: 1) Selecting the root node: Based on the size of the information gain, select the feature with the largest information gain as the root node of the decision tree. This feature can minimize the uncertainty of the data set. 2) Recursive partitioning: Recursively divide each child node and select the feature with the largest information gain as the partitioning criterion until the stopping condition is met. The stopping condition may include: all samples in the node belong to the same category (pure node). The preset maximum tree depth is reached. The number of samples in the node is less than the preset minimum number of samples. 3) Forming a decision tree: Through the above process, a decision tree is constructed. Each branch of the decision tree represents a decision rule (such as "insurance rate > 0.3" or "credit score < 600"), and the leaf node represents the final decision result (insurance or not).

[0059] Step S207: performing risk assessment on the target data based on the target decision tree to obtain a corresponding risk assessment result.

[0060] In this embodiment, the risk assessment process includes: applying a decision tree: performing a risk assessment on the input target data based on the constructed target decision tree. Starting from the root node, go down along the branches of the decision tree according to the characteristic values ​​of the data. For example, if the customer's risk rate is 0.4, go down along the branch of "risk rate>0.3"; if the credit score is 550, go down along the branch of "credit score<600". Arrive at the leaf node: finally reach the leaf node and get the risk assessment result. Among them, the risk assessment result can be the user's risk level (which may include low risk, medium risk or high risk), and the specific generation process of the user's risk level will be further described in detail in the subsequent specific embodiments of this application, so it will not be elaborated here.

[0061] A corresponding underwriting decision result is generated based on the risk assessment result, and the underwriting decision result is sent to the user.

[0062] In this embodiment, the above-mentioned risk assessment result may refer to the risk level of the user, and based on the association between the risk level and the underwriting decision, a corresponding underwriting decision result may be generated based on the risk assessment result.

[0063] Specifically, the relationship between risk level and underwriting decisions includes the following: 1) Low-risk customers: Underwriting likelihood: High. Low-risk customers typically have a good credit history, a low accident rate, and a stable occupation and income. These characteristics indicate that these customers are less risky, making insurance companies more willing to insure them. Premium level: Low-risk customers' premiums are typically at the lower end of the basic range, and they may even receive certain premium discounts. This is because their risk is lower, and the insurance company bears less claims risk. 2) Medium-risk customers: Underwriting likelihood: Medium. Medium-risk customers may have certain risk factors based on certain characteristics, but their overall risk remains manageable. Insurance companies will assess each case and decide whether to insure them. Premium level: Medium-risk customers' premiums are typically in the middle of the basic range. Insurance companies may increase premiums to cover potential risks, but generally will not deny coverage. 3) High-risk customers: Underwriting likelihood: Low. High-risk customers may have a high accident rate, a poor credit history, or work in a high-risk occupation. These characteristics indicate that these customers are more risky, and insurance companies may be cautious about insuring them. Underwriting conditions: If an insurance company decides to insure a high-risk customer, it may impose certain conditions, such as increasing premiums, increasing deductibles, or limiting coverage, to reduce its own claims risk. Denial of coverage: In some cases, if a customer's risk is too high, the insurance company may directly refuse coverage to avoid potentially significant claims losses. Underwriting decisions are made based on the risk level, including whether to insure, the premium level, and the underwriting conditions.

[0064] In addition, if it is determined that the user is eligible for insurance, an underwriting decision result indicating that the user is eligible for insurance will be generated and sent to the user. The system will also calculate the user's premium range based on the weights of each feature in the decision-making process (such as risk rate, credit score, insurance amount, etc.). The calculation of the premium range usually takes into account the risk level and market competition. The system will then feed back the underwriting decision result and the premium range to the user, and the user can choose a suitable premium within the range for insurance. For example, the system may recommend a premium range of "1,000 yuan to 1,500 yuan", and the customer can choose 1,200 yuan as the final premium. After the user confirms the premium, the insurance process ends, and the system generates the policy and notifies the user.

[0065] In addition, if it is judged that the user cannot be insured, the system generates an uninsurable insurance decision result. The feedback information will explain the reason for the uninsurability in detail, for example: the risk rate is too high (such as the number of risks in the past three years is more than 3 times). The credit is low (such as the credit score is lower than 600). The occupation risk is too high (such as engaging in high-risk occupations such as race car driver). After the user receives the feedback of uninsurability, he can choose to adjust the insurance plan (such as increasing the self-paid amount, selecting other insurance products) or find other insurance companies.

[0066] The application first receives a user-triggered insurance processing request; wherein the insurance processing request carries the basic information of the user; then extracts the basic information from the insurance processing request, and judges whether the user is a renewal user based on the basic information; if so, queries the related data of the user during the historical insurance period from the preset database; then constructs the portrait data of the user based on the basic information and the related data; and formats the portrait data to obtain the corresponding target data; subsequently, the target data is subjected to information gain calculation processing based on a preset decision model, and a target decision tree is constructed based on the obtained target information gain; further, the target data is subjected to risk assessment based on the target decision tree, to obtain a corresponding risk assessment result; finally, a corresponding insurance decision result is generated based on the risk assessment result, and the insurance decision result is sent to the user. Based on the above intelligent underwriting process, the application applies a decision model to automatically underwrite the insurance processing request triggered by the user, significantly reducing the time and cost of manual review, reducing human intervention, and effectively improving the underwriting efficiency. Moreover, the application of the decision model makes the underwriting result more objective and transparent, and each step of decision making is supported by data, thereby improving the accuracy and credibility of the underwriting result.

[0067] In some optional implementations, step S204 includes the following steps:

[0068] Obtain investigation data corresponding to the user.

[0069] In the embodiment, the investigation data mentioned above refers to first-hand data of the insurance subject obtained by the business personnel through on-site investigation of the user. For example, for property insurance, the actual condition of the property and the safety of the surrounding environment are investigated; for life insurance, the health status and living environment of the client are understood. The business personnel enter the collected investigation data into the system to provide data support for subsequent risk assessment.

[0070] Obtain credit information of the user.

[0071] In the embodiment, the credit information refers to the bank credit information of the user, and the accuracy of the credit data can be ensured by querying the credit report or combining with the credit certificate provided by the client.

[0072] The basic information, the investigation data, the credit information, and the related data are integrated to obtain corresponding integrated data.

[0073] In the embodiment, the basic information, the investigation data, the credit data, and the related data are integrated, and the integrated data is used as the portrait data of the user.

[0074] The integrated data is used as the portrait data of the user.

[0075] In the embodiment, the data collection is a crucial link in the underwriting process, and directly relates to the accuracy of the subsequent risk assessment and the effectiveness of the decision.

[0076] The application obtains the investigation data corresponding to the user, and obtains the credit information of the user. Then, the basic information, the investigation data, the credit information, and the related data are integrated to obtain corresponding integrated data. The integrated data is used as the portrait data of the user. Based on the above processing procedure, the application collects the investigation data, the credit information, the basic information of the user, and the related data of the user during the historical underwriting from multiple dimensions, and combines all the information to generate the portrait data of the user, effectively ensuring the accuracy and comprehensiveness of the generated portrait data. The accuracy of the risk assessment and the effectiveness of the underwriting decision can be effectively improved by using the subsequent risk assessment based on the portrait data.

[0077] In some optional implementation manners of the embodiment, step S205 includes the following steps:

[0078] The numerical features in the related data are standardized to obtain corresponding first feature data.

[0079] In this embodiment, the standardization of the numerical features includes: for numerical features such as claim rate, target age (such as vehicle usage age), policyholder age, credit score, etc., the system will perform standardization. The purpose of standardization is to eliminate the influence of different dimensions and make the data on a unified scale for comparison and analysis. Specifically, claim rate standardization includes: scaling the claim rate (such as the ratio of the number of claims in the past three years to the total mileage) according to a certain range, for example, using the min-max standardization method, mapping it to the [0, 1] interval. This can be achieved by calculating (current value-min value) / (max value-min value). Age standardization includes: similarly, standardize the policyholder's age so that it also falls within the [0, 1] interval. For example, if the age range is 18 to 80 years old, then 40 years old can be standardized to (40-18) / (80-18)≈0.34. For other numerical features: similar methods can be applied to other numerical features such as credit score, insurance amount, etc. to ensure that they are analyzed on the same scale.

[0080] The category features in the related data are encoded to obtain corresponding second feature data.

[0081] In this embodiment, the encoding of the category features includes: for category features such as the use environment of the target (city, countryside), the occupation of the policyholder (engineer, teacher, doctor, etc.), the insurance type (vehicle insurance, life insurance, property insurance), etc. The system will perform one-hot encoding. One-hot encoding converts each category into a binary vector so that the model can process these non-numerical data. Specifically, use environment encoding: for example, "city" is encoded as [1, 0], and "countryside" is encoded as [0, 1]. In this way, the model can more easily identify the differences between different categories. Occupation encoding: For the occupation of the policyholder, it can be encoded into a longer binary vector, each occupation corresponds to a unique combination. For example, engineer is encoded as [1, 0, 0], teacher is encoded as [0, 1, 0], and doctor is encoded as [0, 0, 1], etc. For category features: similar methods can be applied to other category features such as insurance type, gender, education level, etc. to ensure that they are input into the model in numerical form.

[0082] The first feature data and the second feature data are merged to obtain corresponding merged feature data.

[0083] In the embodiment, the merging processing includes collating and packing the first feature data and the second feature data according to the input format required by the decision model. Specifically, it includes merging the numerical features (first feature data) and the category features (second feature data) into a complete data set, and ensuring that the position and format of each feature are consistent with the expected input of the model, so as to obtain the corresponding merged feature data.

[0084] The merged feature data is taken as the target data.

[0085] In the embodiment, the formatting processing is a key step of converting the original data into a form suitable for processing by the decision model.

[0086] The application obtains the corresponding first feature data by standardizing the numerical features in the related data, then obtains the corresponding second feature data by encoding the category features in the related data, then merges the first feature data and the second feature data to obtain the corresponding merged feature data, and finally takes the merged feature data as the target data. Based on the above processing flow, the system can eliminate the dimensional difference and category difference between the data by standardizing and one-hot encoding the portrait data, so that the decision model can more accurately learn the rules in the data. Furthermore, inputting the formatted target data into the decision model can provide a basis for subsequent model analysis and decision. The successful implementation of this step not only improves the accuracy and processing efficiency of the data evaluation processing of the decision model, but also provides strong technical support for the risk assessment and decision of the insurance company.

[0087] In some optional implementations, step S207 includes the following steps:

[0088] Based on the target decision tree, the target data is subjected to risk score calculation processing to obtain a corresponding comprehensive risk score.

[0089] In this embodiment, the above risk score calculation process includes: feature weight determination: by reviewing the target decision tree, the weight of each feature in the decision-making process is determined. These weights reflect the degree of influence of different features on the final decision. For example, the impact of the claim rate on the risk score may be greater, while the impact of the credit score is smaller. Feature value standardization: in order to ensure the comparability of different features in the score, the specific feature value of the user is standardized. For example, the feature values of claim rate, credit score, age, etc. are mapped to a unified scale (such as between 0 and 1). Weighted summation: according to the weight of the feature and the standardized feature value, a comprehensive risk score is calculated. For example, the comprehensive risk score = (claim rate weight x claim rate standardized value) + (credit score weight x credit score standardized value)... In this way, customers with high claim rates and low credit scores will have a higher risk score.

[0090] Obtain a preset risk level division strategy.

[0091] In this embodiment, the strategy content of the above risk level division strategy includes: score interval setting: the system will set different score intervals to divide the risk level according to historical data and business needs. For example, the following intervals can be set: low risk: the comprehensive risk score is between 0 and 0.3. Medium risk: the comprehensive risk score is between 0.3 and 0.6. High risk: the comprehensive risk score is between 0.6 and 1.0. Among them, the division of risk levels can be adjusted according to actual business needs. For example, if the company wants to attract more low-risk customers, it can appropriately lower the premium range of the corresponding low-risk customers.

[0092] Determine the target risk level corresponding to the comprehensive risk score based on the risk level division strategy.

[0093] In this embodiment, the target risk level corresponding to the comprehensive risk score can be determined based on the strategy content of the above risk level division strategy.

[0094] Take the target risk level as the risk assessment result corresponding to the target data.

[0095] The application obtains a corresponding comprehensive risk score through risk score calculation processing of the target data based on the target decision tree; then acquires a preset risk level division strategy; then determines a target risk level corresponding to the comprehensive risk score based on the risk level division strategy; and subsequently takes the target risk level as a risk assessment result corresponding to the target data. Based on the above processing procedure, the application accurately calculates a comprehensive risk score corresponding to target data based on the use of a target decision tree, and then divides users into different risk levels according to the comprehensive risk score based on the use of a risk level division strategy, thereby automatically and accurately completing risk assessment of target data, improving the processing efficiency of risk assessment, and ensuring the accuracy of the obtained risk assessment result. This process not only helps insurance companies to quantify the risk level of customers, but also provides a scientific basis for premium pricing. By reasonably dividing risk levels, the insurance company can provide personalized insurance products and services for users under the premise of controllable risk.

[0096] In some optional implementations, after step S208, the electronic device can further perform the following steps:

[0097] It is determined whether the underwriting decision result is insurable.

[0098] In this embodiment, the content of the underwriting decision result includes insurable or uninsurable. The content of the underwriting decision result can be analyzed to determine whether the underwriting decision result is insurable.

[0099] If yes, a corresponding basic premium is calculated based on the risk assessment result.

[0100] In this embodiment, the risk assessment result refers to the target risk level of the user. Each risk level corresponds to a basic range of premium. For example, the basic premium range of low-risk customers can be 1000-1500 yuan. The basic premium range of medium-risk customers can be 1500-2000 yuan. The basic premium range of high-risk customers can be 2000-2500 yuan. In addition, the division of risk levels and the corresponding premium range can be adjusted according to actual business needs. For example, if the company wants to attract more low-risk customers, the premium range of corresponding low-risk customers can be appropriately reduced.

[0101] For example, the basic premium of low-risk customers can be 1000 yuan, the basic premium of medium-risk customers can be 1500 yuan, and the basic premium of high-risk customers can be 2000 yuan.

[0102] A preset market adjustment factor is acquired.

[0103] In this embodiment, the system understands the premium level and market competition of similar insurance products through regular market research. This includes analyzing the pricing strategies, product characteristics, and customer feedback of competitors. Then, according to the market research results, the system will determine a market adjustment factor to adjust the premium interval to ensure that the company's premium level is competitive in the market.

[0104] The base premium is adjusted based on the market adjustment factor to obtain corresponding target premium data.

[0105] In this embodiment, the adjustment of the base premium includes multiplying the base premium by the market adjustment factor to obtain the final premium interval, which is used as the target premium data. For example, if the market adjustment factor is 1.2, the premium interval for low-risk customers may be 1000-1200 yuan. The premium interval is a range, not a fixed value, to provide customers with some selection space. For example, the system may suggest a premium interval of "1000-1500 yuan", and the customer can choose the appropriate premium within this range.

[0106] The target premium data is sent to the user.

[0107] In this embodiment, after sending the target premium data to the user, the system also collects user feedback on the target premium data to understand the customer's acceptance and selection preference of the premium. Then, according to customer feedback and market changes, the system dynamically adjusts the calculation method of the premium interval and the market adjustment factor to ensure the reasonableness and competitiveness of the premium level.

[0108] The application determines whether the underwriting decision result is insurable, calculates the corresponding base premium based on the risk assessment result if it is, obtains a preset market adjustment factor, adjusts the base premium based on the market adjustment factor to obtain corresponding target premium data, and then sends the target premium data to the user. Based on the above processing flow, when the underwriting decision result is detected as insurable, the application automatically calculates the corresponding base premium based on the risk assessment result, then adjusts the base premium based on the market adjustment factor to obtain the corresponding target premium data, and sends the target premium data to the user. In this way, the application calculates the base premium based on the risk assessment result, and then adjusts the base premium based on the market adjustment factor, so as to intelligently and accurately provide a reasonable premium interval for the user. This process not only ensures the fairness and reasonableness of the premium, but also enhances the user's acceptance and satisfaction of the insurance product. The dynamic adjustment mechanism enables the premium interval to adapt to market changes and changes in user demand, maintaining the company's competitive advantage.

[0109] In some optional implementations of the embodiment, after step S206, the electronic device can further perform the following steps:

[0110] Obtaining pre-collected specified customer data.

[0111] In the embodiment, the specified customer data refers to all customer data accumulated in relation to the underwriting. Specifically, during the long-term operation of the system, the historical data of the renewal user will be updated continuously. For example, the customer's loss record, credit score, occupation information, etc. may change over time, and the system will update these data regularly to maintain their timeliness. Moreover, as new customers continue to join, the system will accumulate a large amount of new customer insurance data. These data include the customer's basic information, insurance subject details, credit information, etc., providing rich materials for model optimization. And all accumulated data will be properly stored and managed, usually stored in a database or data warehouse, for subsequent analysis and model training.

[0112] Based on the preset evaluation indicators, the specified customer data is used to evaluate the decision model to obtain a corresponding model evaluation result.

[0113] In the embodiment, the selection of the above evaluation indicators is not specifically limited and can be determined according to actual business needs, for example, can include accuracy (the proportion of correct predictions by the model), recall rate (the proportion of positive examples correctly identified by the model), F1 value (the harmonic mean of accuracy and recall rate), etc. These indicators can comprehensively reflect the performance of the model.

[0114] Among them, the decision model can be evaluated regularly according to the selected evaluation indicators, by comparing the prediction results of the decision model with the actual business results (such as the final underwriting decision, the actual loss situation of the customer, etc.), to understand the performance and problems of the decision model in actual application. Further, through model evaluation, possible biases or deficiencies of the decision model are identified, such as inaccurate prediction of certain customer groups, insufficient adaptability to new data types, etc., and corresponding model evaluation results are generated.

[0115] Based on the model evaluation result, a corresponding model optimization strategy is determined.

[0116] In the embodiment, the strategy content of the model optimization strategy includes: adjusting the feature selection for model training according to the model evaluation result. For example, increasing or decreasing certain features, or re-evaluating the importance of the features. And modifying the construction rules of the decision tree, such as adjusting the threshold of information gain, changing the splitting standard of the tree, increasing or decreasing the maximum depth of the tree, etc., to improve the generalization ability and accuracy of the model. And using the updated data and adjusted parameters to retrain the decision tree model to generate a new model version.

[0117] Adjust and optimize the decision model based on the model optimization strategy.

[0118] In this embodiment, the decision model can be adjusted and optimized based on the strategy content of the above-mentioned model optimization strategy. Subsequently, the optimized model can be applied to the actual underwriting process to replace the original model version. The new model can better adapt to changes in business and updates in data, improving the accuracy and efficiency of underwriting. Further, feedback information on the model output can be collected, such as the customer's acceptance of the premium interval and the understanding of the non-insurance reasons. These feedbacks can help further improve the model. Further, based on the user feedback and the actual performance of the model, the model is continuously adjusted and optimized, forming a virtuous cycle. This process ensures that the model is always in the best state and can provide the best service to customers.

[0119] The present application obtains pre-collected specified customer data, then uses the specified customer data to evaluate the decision model based on the preset evaluation index to obtain the corresponding model evaluation result, then determines the corresponding model optimization strategy based on the model evaluation result, and subsequently adjusts and optimizes the decision model based on the model optimization strategy. Based on the above processing flow, the present application obtains pre-collected specified customer data, and uses the specified customer data to evaluate the decision model based on the preset evaluation index to obtain the model evaluation result, so as to timely find out the problems of the decision model, and then adjust and optimize the decision model based on the model optimization strategy corresponding to the model evaluation result, so that the subsequent optimized decision model can effectively improve the quality and efficiency of underwriting and enhance the competitiveness of the insurance company.

[0120] In some optional implementation manners of the present embodiment, after step S204, the electronic device can further perform the following steps:

[0121] Obtain the data type of the portrait data.

[0122] In this embodiment, the portrait data can be type-extracted to obtain the data type corresponding to the portrait data. The data type can include structured data or unstructured data.

[0123] Determine the corresponding specified storage medium based on the data type.

[0124] In the embodiment, the appropriate database can be selected as the specified storage medium according to the data type and access requirement of the portrait data. The relational database is suitable for structured data and facilitates complex query and correlation analysis; the non-relational database is suitable for processing a large amount of unstructured data or semi-structured data.

[0125] The portrait data is classified and labeled to obtain corresponding target portrait data.

[0126] In the embodiment, the classification and labeling process includes classifying and labeling the data when saving the data, so as to facilitate subsequent query and analysis. For example, the basic information, insurance information, accident record, credit information and the like of the customer are respectively stored in different tables, and the data is associated through the foreign key.

[0127] The target portrait data is stored in the specified storage medium.

[0128] In the embodiment, after the data storage of the target portrait data based on the use of the specified storage medium is completed, the security measures (such as encryption and access control) are further used to prevent data leakage or damage, so as to ensure the security and integrity of the target portrait data. Meanwhile, the target portrait data can be backed up regularly to prevent data loss.

[0129] The application obtains the data type of the portrait data, determines the corresponding specified storage medium based on the data type, classifies and labels the portrait data to obtain the corresponding target portrait data, and then stores the target portrait data in the specified storage medium. Based on the above processing flow, the application determines the corresponding specified storage medium based on the data type of the portrait data, classifies and labels the portrait data to obtain the target portrait data, and then stores the target portrait data in the specified storage medium, so as to realize scientific data saving and management of the target portrait data, so that the insurance company can ensure the traceability and analyzability of the target portrait data, thereby providing strong support for future business decision and optimization.

[0130] In some optional implementations, the obtained user information seeks the consent of the user and meets the requirements of relevant laws and relevant policies.

[0131] In addition, the non-company software tools or components appearing in the embodiments of the application are only examples and do not represent actual use.

[0132] It should be understood that the size of the serial number of each step in the above embodiments does not mean the execution order, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the application.

[0133] It should be emphasized that, in order to further ensure the privacy and security of the above risk assessment result, the above risk assessment result can also be stored in a node of a block chain.

[0134] The embodiments of the present application can acquire and process related data based on artificial intelligence technology. Artificial intelligence (AI) is the use of digital computers or computer-controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0135] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by computer readable instructions instructing related hardware, and the computer readable instructions can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiments of each method. Among them, the storage medium can be a non-volatile storage medium such as a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0136] It should be understood that although each step in the flowchart of the accompanying drawings is displayed in sequence according to the arrow, these steps are not necessarily executed in sequence according to the arrow. Unless otherwise stated herein, the execution of these steps is not strictly limited in sequence, and they can be executed in other orders. Moreover, at least one of the steps in the flowchart of the accompanying drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order is not necessarily sequential, but can be alternately or alternately executed with at least one part of other steps or other steps. Sub-steps or stages.

[0137] Further referring to Figure 3 , as an implementation of the method shown in Figure 2 , the present application provides an embodiment of a data evaluation device, which corresponds to the method embodiment shown in Figure 2 , and the device can be specifically applied to various electronic devices.

[0138] As shown in Figure 3 , the data evaluation device 300 described in the embodiment includes a receiving module 301, a first judgment module 302, a query module 303, a construction module 304, a first processing module 305, a second processing module 306, a first evaluation module 307, and a third processing module 308. Among them:

[0139] The receiving module 301 is configured to receive a user triggered insurance application processing request; wherein the insurance application processing request carries basic information of the user;

[0140] The first determining module 302 is configured to extract the basic information from the insurance application processing request, and determine whether the user is a renewal user based on the basic information;

[0141] The querying module 303 is configured to, if yes, query related data of the user during a historical underwriting period from a preset database;

[0142] The constructing module 304 is configured to construct portrait data of the user based on the basic information and the related data;

[0143] The first processing module 305 is configured to perform format processing on the portrait data to obtain corresponding target data;

[0144] The second processing module 306 is configured to perform information gain calculation processing on the target data based on a preset decision model, and construct a corresponding target decision tree based on obtained target information gain;

[0145] The first evaluating module 307 is configured to perform risk evaluation on the target data based on the target decision tree, and obtain a corresponding risk evaluation result;

[0146] The third processing module 308 is configured to generate a corresponding underwriting decision result based on the risk evaluation result, and send the underwriting decision result to the user.

[0147] In some optional implementation manners of the embodiment, the constructing module 304 includes:

[0148] The first obtaining sub-module is configured to obtain investigation data corresponding to the user;

[0149] The second obtaining sub-module is configured to obtain credit information of the user;

[0150] The integrating sub-module is configured to perform integration processing on the basic information, the investigation data, the credit information, and the related data to obtain corresponding integrated data;

[0151] The first determining sub-module is configured to take the integrated data as the portrait data of the user.

[0152] In some optional implementation manners of the embodiment, the first processing module 305 includes:

[0153] The first processing sub-module is configured to perform standardization processing on numerical features in the related data to obtain corresponding first feature data;

[0154] a second processing submodule, configured to perform encoding processing on the category features in the related data to obtain corresponding second feature data;

[0155] a merging submodule, configured to perform merging processing on the first feature data and the second feature data to obtain corresponding merged feature data;

[0156] a second determining submodule, configured to take the merged feature data as the target data.

[0157] In some optional implementations of the embodiment, the first evaluation module 307 includes:

[0158] a calculation submodule, configured to perform risk score calculation processing on the target data based on the target decision tree to obtain corresponding comprehensive risk score;

[0159] a third obtaining submodule, configured to obtain a preset risk level division strategy;

[0160] a third determining submodule, configured to determine a target risk level corresponding to the comprehensive risk score based on the risk level division strategy;

[0161] a fourth determining submodule, configured to take the target risk level as a risk evaluation result corresponding to the target data.

[0162] In some optional implementations of the embodiment, the data evaluation apparatus further includes:

[0163] a second judging module, configured to judge whether the underwriting decision result is insurable;

[0164] a calculation module, configured to, if yes, calculate corresponding basic premium based on the risk evaluation result;

[0165] a first obtaining module, configured to obtain a preset market adjustment factor;

[0166] an adjustment module, configured to perform adjustment processing on the basic premium based on the market adjustment factor to obtain corresponding target premium data;

[0167] a sending module, configured to send the target premium data to the user.

[0168] In some optional implementations of the embodiment, the data evaluation apparatus further includes:

[0169] a second obtaining module, configured to obtain pre-collected specified customer data;

[0170] a second evaluation module, configured to perform evaluation processing on the decision model using the specified customer data based on a preset evaluation index to obtain corresponding model evaluation result.

[0171] a first determining module, configured to determine a corresponding model optimization strategy based on the model evaluation result;

[0172] an optimization module, configured to perform corresponding adjustment and optimization processing on the decision model based on the model optimization strategy.

[0173] In some optional implementation manners of the embodiment, the data evaluation apparatus further includes:

[0174] a third obtaining module, configured to obtain a data type of the portrait data;

[0175] a second determining module, configured to determine a corresponding specified storage medium based on the data type;

[0176] a fourth processing module, configured to perform classification and labeling processing on the portrait data to obtain corresponding target portrait data;

[0177] a storage module, configured to store the target portrait data into the specified storage medium.

[0178] To solve the above technical problems, the embodiment of the present application further provides a computer device. For details, please refer to Figure 4 , Figure 4 The basic structure block diagram of the computer device of the embodiment is shown in FIG. 4.

[0179] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 which are connected to each other through a system bus. It should be noted that only the computer device 4 with components 41-43 is shown in the figure, but it should be understood that it is not required to implement all the shown components, and more or fewer components can be alternatively implemented. Among them, those skilled in the art can understand that the computer device here is a device capable of automatically performing numerical calculation and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0180] The computer device can be a desktop computer, a notebook computer, a palm computer, a cloud server, and other computing devices. The computer device can interact with the user through a keyboard, a mouse, a remote controller, a touchpad, a voice control device, and other ways.

[0181] The memory 41 includes at least one type of readable storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or a memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 4. Of course, the memory 41 can also include both an internal storage unit and an external storage device of the computer device 4. In this embodiment, the memory 41 is generally used to store an operating system and various application software installed on the computer device 4, such as computer readable instructions of the data evaluation method, etc. In addition, the memory 41 can also be used to temporarily store various data that have been output or will be output.

[0182] The processor 42 can be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer readable instructions or process data stored in the memory 41, such as computer readable instructions of the data evaluation method.

[0183] The network interface 43 can include a wireless network interface or a wired network interface, and is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0184] The present application also provides another embodiment, i.e., to provide a computer readable storage medium storing computer readable instructions, which can be executed by at least one processor to make the at least one processor perform the steps of the data evaluation method as described above.

[0185] Those skilled in the art can clearly understand the above-mentioned embodiment method can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases, the former is a better embodiment. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes a plurality of instructions for making a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) execute the method described in each embodiment of the present application.

[0186] Obviously, the above-described embodiments are only some of the embodiments of the present application, not all the embodiments, and the drawings show the preferred embodiments of the present application, but do not limit the patent scope of the present application. The present application can be implemented in many different forms, and conversely, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing specific embodiments, or make equivalent replacements to some of the technical features. Any equivalent structure made by using the content of the specification and drawings, directly or indirectly applied to other related technical fields, is also within the scope of the patent protection of the present application.

Claims

1. A data evaluation method, characterized by, The method comprises the following steps: receiving a user triggered insurance application processing request; wherein the insurance application processing request carries basic information of the user; extracting the basic information from the insurance application processing request, and judging whether the user is a renewal user based on the basic information; if yes, querying the related data of the user during the historical underwriting period from a preset database; constructing portrait data of the user based on the basic information and the related data; formatting the portrait data to obtain corresponding target data; performing information gain calculation processing on the target data based on a preset decision model, and constructing a corresponding target decision tree based on the obtained target information gain; performing risk assessment on the target data based on the target decision tree to obtain a corresponding risk assessment result; generating a corresponding underwriting decision result based on the risk assessment result, and sending the underwriting decision result to the user.

2. The data evaluation method according to claim 1, characterized in that The step of constructing the portrait data of the user based on the basic information and the related data comprises: obtaining investigation data corresponding to the user; obtaining credit information of the user; integrating the basic information, the investigation data, the credit information and the related data to obtain integrated data; taking the integrated data as the portrait data of the user.

3. The data evaluation method of claim 1, wherein, The step of formatting the portrait data to obtain corresponding target data comprises: standardizing numerical features in the related data to obtain first feature data; encoding category features in the related data to obtain second feature data; merging the first feature data and the second feature data to obtain merged feature data; taking the merged feature data as the target data.

4. The data evaluation method of claim 1, wherein, The step of performing risk assessment on the target data based on the target decision tree to obtain a corresponding risk assessment result comprises: performing risk score calculation processing on the target data based on the target decision tree to obtain a corresponding comprehensive risk score; obtaining a preset risk level division strategy; determining a target risk level corresponding to the comprehensive risk score based on the risk level division strategy; taking the target risk level as the risk assessment result corresponding to the target data.

5. The data evaluation method of claim 1, wherein, After the step of generating a corresponding underwriting decision result based on the risk assessment result, and sending the underwriting decision result to the user, the method further comprises: judging whether the underwriting decision result is insurable; if yes, calculating a corresponding basic premium based on the risk assessment result; obtaining a preset market adjustment factor; adjusting the basic premium based on the market adjustment factor to obtain corresponding target premium data; sending the target premium data to the user.

6. The data evaluation method of claim 1, wherein, After the step of performing information gain calculation processing on the target data based on a preset decision model, and constructing a corresponding target decision tree based on the obtained target information gain, the method further comprises: obtaining pre-collected specified customer data; Based on the preset evaluation index, the specified customer data is used to evaluate the decision model, and the corresponding model evaluation result is obtained; Based on the model evaluation result, the corresponding model optimization strategy is determined; Based on the model optimization strategy, the decision model is adjusted and optimized accordingly.

7. The data evaluation method of claim 1, wherein, After the step of constructing the user's portrait data based on the basic information and the related data, the method further includes: Obtaining the data type of the portrait data; Based on the data type, the corresponding specified storage medium is determined; Classifying and labeling the portrait data to obtain the corresponding target portrait data; The target portrait data is stored in the specified storage medium.

8. A data evaluation device, characterized by It includes: The receiving module is used for receiving the user triggered insurance processing request; wherein the insurance processing request carries the basic information of the user; The first judgment module is used for extracting the basic information from the insurance processing request, and judging whether the user is a renewal user based on the basic information; If so, the query module queries the related data of the user during the historical underwriting period from the preset database; The construction module is used for constructing the user's portrait data based on the basic information and the related data; The first processing module is used for formatting the portrait data to obtain the corresponding target data; The second processing module is used for calculating the information gain of the target data based on the preset decision model, and constructing the corresponding target decision tree based on the obtained target information gain; The first evaluation module is used for risk assessment of the target data based on the target decision tree, and the corresponding risk assessment result is obtained; The third processing module is used for generating the corresponding underwriting decision result based on the risk assessment result, and sending the underwriting decision result to the user.

9. A computer device, comprising: The memory and the processor, the memory stores computer readable instructions, the processor executes the computer readable instructions to realize the steps of the data evaluation method in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer readable instructions, and the computer readable instructions are executed by the processor to realize the steps of the data evaluation method in any one of claims 1 to 7.

Images