Customer portrait analysis method and system for financial service

Abstract

The embodiment of the invention provides a customer portrait analysis method and system for financial service, and belongs to the technical field of user portraits. The method comprises the following steps: in response to authorization information of a target customer, collecting business related information of the target customer; preprocessing the business related information, and constructing a customer portrait corresponding to the target customer based on the preprocessed business related information; calling a trained risk assessment model based on the customer portrait to obtain a risk assessment result of the corresponding target customer; and based on the customer portrait and the risk assessment result, obtaining a recommendation service of the target customer, and pushing the recommendation service to a target client. According to the scheme of the invention, the accuracy and response efficiency of financial services are improved, the risk control capability and customer experience of the system are enhanced, and an automatic and intelligent business recommendation process is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of user portraits, and in particular to a customer portrait analysis method for financial services and a customer portrait analysis system for financial services. Background Art

[0002] With the rapid development of the energy market, supply chain financial services are increasingly used in the energy industry. Energy companies often involve a large number of asset transactions and capital flows, and financial institutions need to have a comprehensive understanding of the credit status of these companies in order to provide personalized financial services and effective risk control. However, when building customer portraits, existing financial service systems usually have problems such as incomplete data collection, single analysis dimensions, and inaccurate risk identification. This makes it difficult for the system to fully meet customer needs when recommending financial products, and also increases the risks faced by financial institutions.

[0003] Traditional customer profiling systems mainly rely on basic customer information and historical transaction records, and lack in-depth analysis of key dimensions such as customer source channels, shareholder relationships, foreign investments, and financial status. Due to the limitations of information sources and dimensions, financial institutions find it difficult to obtain a comprehensive risk profile when conducting risk assessments, and lack sufficient knowledge of factors that may affect customer credit risk and market risk. Therefore, recommended financial products often fail to accurately match customers' actual needs and risk tolerance, resulting in poor customer experience and insufficient personalization of financial services.

[0004] In addition, the existing system lacks intelligent support in data processing and analysis, and fails to fully utilize advanced technologies such as big data and machine learning, resulting in low accuracy in the analysis of customer behavior patterns and inaccurate predictions of customer preferences. For the dynamically changing energy industry, if financial institutions cannot obtain and analyze the latest customer data in real time, it will be difficult to adjust product strategies in a timely manner, thus affecting the effectiveness and competitiveness of financial services.

[0005] To sum up, the existing financial service recommendation solutions based on customer portraits have problems such as insufficient data dimensions, incomplete risk identification and insufficient intelligent analysis. There is an urgent need for an intelligent analysis solution based on multi-dimensional customer portraits. Summary of the invention

[0006] The purpose of the embodiments of the present invention is to provide a customer portrait analysis method and system for financial services, so as to at least solve the problems of insufficient data dimensions, incomplete risk identification and insufficient intelligent analysis in existing financial service recommendation solutions based on customer portraits.

[0007] In order to achieve the above-mentioned objectives, the first aspect of the present invention provides a customer portrait analysis method for financial services, the method comprising: in response to the authorization information of the target customer, collecting business-related information of the target customer; performing preprocessing on the business-related information, and constructing a customer portrait of the corresponding target customer based on the preprocessed business-related information; calling a trained risk assessment model based on the customer portrait to obtain a risk assessment result of the corresponding target customer; obtaining recommended business for the target customer based on the customer portrait and the risk assessment result, and pushing the recommended business to the target client.

[0008] Optionally, in response to the target customer's authorization information, collecting the target customer's business-related information includes: receiving and verifying the target customer's authorization information based on the SpringBoot framework; after the verification is passed, calling each microservice based on the SpringCloud distributed architecture to collect the customer's business-related information in real time from multiple data sources; temporarily storing the collected business-related information in the Redis cache, and writing the cached data to the PostgreSQL database at predetermined intervals.

[0009] Optionally, the preprocessing of the business-related information includes: cleaning the business-related information based on the SpringBoot framework; converting the format of the cleaned business-related information into business-related information in a unified preset target format; and performing standardization processing on the business-related information after the format conversion to obtain preprocessed business-related information.

[0010] Optionally, a customer profile of the corresponding target customer is constructed based on the preprocessed business-related information, including: classifying the preprocessed business-related information based on predetermined portrait indicators to obtain a data set corresponding to each portrait indicator; constructing corresponding portrait indicators based on each data set; and combining the portrait indicators to obtain a customer profile of the target customer.

[0011] Optionally, the portrait indicators include: any one or more of customer basic information indicators, shareholder relationship indicators, foreign investment indicators, financial status indicators and historical risk assessment indicators.

[0012] Optionally, the method also includes: performing risk assessment model training; the training rules of the risk assessment model are: collecting historical business data and performing preprocessing on the historical business data; classifying the preprocessed historical business data based on each risk indicator, obtaining a data set corresponding to each risk indicator, and performing risk level labeling on each data set; constructing training samples based on the labeled data sets, performing model training based on the training samples, and obtaining a risk assessment model.

[0013] Optionally, the risk indicators include: any one or more of credit risk, market risk and operational risk.

[0014] Optionally, obtaining recommended services for target customers based on the customer portrait and the risk assessment results, and pushing the recommended services to target clients, includes: obtaining recommended services for target customers based on the customer portrait and the risk assessment results based on collaborative filtering and / or content recommendation algorithms; wherein the recommended services are any one or more of loan business, insurance business and leasing business.

[0015] The second aspect of the present invention provides a customer portrait analysis system for financial services, the system comprising: a collection unit, used to collect business-related information of a target customer in response to the authorization information of the target customer; a processing unit, used to perform preprocessing on the business-related information, and construct a customer profile of the corresponding target customer based on the preprocessed business-related information; a risk assessment unit, used to call a trained risk assessment model based on the customer portrait to obtain a risk assessment result of the corresponding target customer; a business recommendation unit, used to obtain recommended business for the target customer based on the customer portrait and the risk assessment result, and push the recommended business to the target client.

[0016] On the other hand, the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed on a computer, enables the computer to execute the above-mentioned customer profile analysis method for financial services.

[0017] Through the above technical solution, the solution of the present invention collects business-related data by responding to the authorization information of the target customer, and performs preprocessing on the data to ensure the accuracy and standardization of the information. Then, the system builds a customer portrait based on the preprocessed information to fully reflect the customer's business characteristics and behavior patterns. On this basis, the pre-trained risk assessment model is called to generate the risk assessment results of the customer, so as to accurately identify and monitor customer risks. Finally, based on the customer portrait and risk assessment results, the system recommends suitable services to the customer and pushes them to the client to make the recommendations more personalized and targeted. This method improves the accuracy and response efficiency of financial services, enhances the system's risk control capabilities and customer experience, and realizes an automated and intelligent business recommendation process.

[0018] Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0020] Figure 1 It is a flowchart of the steps of a customer portrait analysis method for financial services provided by an embodiment of the present invention;

[0021] Figure 2 It is a system structure diagram of a customer portrait analysis system for financial services provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0022] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0023] Figure 1 1 is a method flow chart of a method for analyzing customer portraits of financial services provided by an embodiment of the present invention. Figure 1 As shown, an embodiment of the present invention provides a method for analyzing customer portraits of financial services, the method comprising:

[0024] Step S10: In response to the authorization information of the target customer, collect the business-related information of the target customer.

[0025] Specifically, the authorization information of the target customer is received and verified based on the SpringBoot framework; after the verification is passed, the SpringCloud distributed architecture calls each microservice to collect the customer's business-related information from multiple data sources in real time; the collected business-related information is temporarily stored in the Redis cache, and the cached data is written to the PostgreSQL database at predetermined intervals.

[0026] In the embodiment of the present invention, after receiving the authorization information of the target customer, authorization verification is performed based on the SpringBoot framework. The SpringBoot framework provides efficient RESTful API support and flexible permission management functions. Through this framework system, the authorization information submitted by the customer can be quickly received, parsed and verified, ensuring that only legally authorized customer information can be collected and processed. This process not only improves security, but also effectively prevents unauthorized data access, ensuring the security of customer privacy and the legality and compliance of data collection.

[0027] After the authorization verification is passed, each microservice is called based on the SpringCloud distributed architecture to collect the customer's business-related information from multiple data sources in real time. SpringCloud provides a flexible and scalable distributed architecture. Through the decoupling characteristics of its microservice modules, it can easily integrate information from different data sources, such as basic customer information, transaction records, historical financial data, etc., and update the customer's business dynamics in real time. This multi-source data real-time collection method can comprehensively cover the customer's business status, thereby providing multi-dimensional basic data for subsequent data analysis and customer portrait construction.

[0028] In order to ensure the real-time and access efficiency of data, the collected customer business information is temporarily stored in the Redis cache. Redis is a high-speed cache database with extremely fast read and write speeds and excellent data persistence capabilities, which can greatly improve the efficiency of data processing. Temporarily storing real-time data in Redis can not only speed up data reading and writing, but also effectively reduce the access pressure of the database and avoid performance bottlenecks caused by frequent access to the database. In addition, the Redis cache can also respond quickly when data is updated or read, providing timely data support for subsequent customer portrait construction and analysis.

[0029] In order to ensure data integrity and long-term storage needs, the data in the cache is written to the PostgreSQL database in batches within a predetermined period. As a relational database, PostgreSQL supports powerful transaction management and data indexing functions, which can ensure data persistence and consistency. This mechanism achieves a balance between the cache and the database, which can not only meet the efficiency requirements of real-time data collection, but also ensure data integrity and secure storage. The periodic data synchronization method effectively avoids the risk of system interruption or cache data loss, and provides a stable data foundation for long-term customer data analysis.

[0030] Based on the solution of the present invention, through the efficient integration of SpringBoot and SpringCloud, combined with the reasonable combination of Redis cache and PostgreSQL database, an efficient and distributed data collection and storage system is built, which not only ensures the real-time and security of customer data collection, but also guarantees the persistent storage requirements of data. This design enhances stability and scalability, provides data guarantee for accurate recommendations of financial services, and also improves the response speed and overall performance in large-scale data processing, significantly improving the accuracy of customer service and user experience.

[0031] Step S20: preprocess the business-related information, and construct a customer profile of the corresponding target customer based on the preprocessed business-related information.

[0032] Specifically, the preprocessing of the business related information includes: cleaning the business related information based on the SpringBoot framework; converting the format of the cleaned business related information to business related information in a unified preset target format; performing standardization processing on the business related information after format conversion to obtain preprocessed business related information.

[0033] Furthermore, a customer profile of the corresponding target customer is constructed based on the preprocessed business-related information, including: classifying the preprocessed business-related information based on predetermined portrait indicators to obtain a data set corresponding to each portrait indicator; constructing a corresponding portrait indicator based on each data set; and obtaining a customer profile of the target customer based on each portrait indicator.

[0034] Specifically, the portrait indicators include: any one or more of customer basic information indicators, shareholder relationship indicators, foreign investment indicators, financial status indicators and historical risk assessment indicators.

[0035] In the embodiment of the present invention, cleaning is the first step in data preprocessing, and its purpose is to remove redundant, missing, outlier and other information in the data that does not meet the analysis requirements. By performing data cleaning based on the SpringBoot framework, incomplete or erroneous data records can be quickly screened and corrected, such as removing duplicate records, filling missing data or correcting format errors. The purpose of cleaning is to improve the accuracy and completeness of the data, which not only ensures the reliability of the data, but also greatly reduces the errors and risks in subsequent analysis.

[0036] Furthermore, in the data processing process after cleaning, the cleaned business information is converted into a unified preset format. The purpose of this step is to unify the format of data from different sources, such as converting different date formats, currency units or encoding formats into the standard format of the system. By converting the data into a unified format through the SpringBoot framework, it can ensure that the subsequent calls and processing of the data by the system will not cause errors due to format differences, and also improve the compatibility between different data sources.

[0037] Furthermore, after the format conversion is completed, the business information needs to be standardized. Standardization refers to normalizing or transforming the data so that the distribution of the data conforms to the expected range, which is convenient for the subsequent processing of machine learning algorithms or data mining algorithms. For example, numerical data is scaled and mapped to a specific interval to prevent extreme values ​​or outliers in the data from deviating from the analysis results. Business information that has been standardized is more suitable for model training and analysis, thus laying a solid data foundation for the subsequent construction of customer portraits.

[0038] Specifically, after the data is preprocessed, it will enter the stage of building a customer portrait. The construction of the customer portrait is based on a preset portrait indicator system, and the preprocessed business information is classified to obtain a data set of various portrait indicators. Business information is classified according to the predetermined portrait indicators, and different types of information are divided into corresponding data sets. For example, the customer's identity information is divided into basic information indicators, the customer's shareholder relationship and external investment records are divided into shareholder relationship and external investment indicators, and financial data is divided into financial status indicators. In addition, historical risk assessment data is also divided independently to provide historical risk background support. This classification process ensures the structured management of various types of data, so that data of different dimensions can be processed and analyzed separately.

[0039] Furthermore, after completing the indicator classification, specific portrait indicators are constructed based on each data set. For example, financial status indicators are constructed based on financial data sets to extract information such as the customer's debt-to-asset ratio and cash flow; based on shareholder relations and external investment data, the customer's shareholder structure and investment distribution can be identified; and historical risk assessment data can reveal the customer's risk level and abnormal behavior patterns in past transactions. By concretizing various types of data into clear portrait indicators, the system can present the customer's business and risk status in a more comprehensive and detailed manner.

[0040] Furthermore, after constructing each independent portrait indicator, the indicators are combined to generate a complete portrait of the target customer. By integrating data from different dimensions such as basic information, shareholder relations, and financial status, the system can comprehensively portray the customer's overall business and risk status. This combined portrait not only shows the customer's current business characteristics, but also provides financial institutions with key indicators such as the customer's financial stability, shareholder structure, and investment preferences, so that financial institutions can comprehensively evaluate the customer's risks and potential value when making decisions.

[0041] Based on the solution of the present invention, high-quality customer data management and accurate portrait generation are achieved. The cleaning, format conversion and standardization in the preprocessing process ensure the accuracy and consistency of the data and avoid the interference of data errors on the analysis results; in the portrait construction stage, the classification and combination of portrait indicators provide rich business and risk dimensions for customer portraits, making customer portraits not only comprehensive but also accurate. Through these methods, the system can provide financial institutions with efficient customer evaluation methods, making business recommendations and risk control more accurate and effective.

[0042] Step S30: calling the trained risk assessment model based on the customer profile to obtain the risk assessment result of the corresponding target customer.

[0043] Specifically, the training rules of the risk assessment model are: collect historical business data and perform preprocessing on the historical business data; classify the preprocessed historical business data based on various risk indicators, obtain the data sets corresponding to each risk indicator, and perform risk level labeling on each data set; construct training samples based on the labeled data sets, perform model training based on the training samples, and obtain a risk assessment model.

[0044] Furthermore, the risk indicators include: any one or more of credit risk, market risk and operational risk.

[0045] In an embodiment of the present invention, a trained risk assessment model is called based on the customer profile to obtain a risk assessment result corresponding to the target customer.

[0046] Specifically, the training rules of the risk assessment model are: collect historical business data and perform preprocessing on the historical business data; classify the preprocessed historical business data based on various risk indicators, obtain the data sets corresponding to each risk indicator, and perform risk level labeling on each data set; construct training samples based on the labeled data sets, perform model training based on the training samples, and obtain a risk assessment model.

[0047] Furthermore, the risk indicators include: any one or more of credit risk, market risk and operational risk.

[0048] In an embodiment of the present invention, in order to ensure the accuracy and reliability of the risk assessment model, a large amount of historical business data needs to be collected when training the model. Historical business data is the basis for building a model. These data can provide the model with rich risk information and help identify risk characteristics in different business activities. After data collection, preprocessing is performed on the historical business data to ensure the accuracy and consistency of the data. For example, outliers and redundant information in the data are removed through data cleaning, and the standardization of the data is improved through format conversion and standardization processing. These processing measures can ensure that the model will not be affected by data noise or errors in subsequent training.

[0049] Furthermore, the preprocessed data is classified according to different risk indicators, which cover multiple dimensions such as credit risk, market risk and operational risk. The purpose of classification is to extract different types of risk information separately, so that the model can focus on specific risk characteristics, thereby improving the accuracy of risk prediction. On this basis, the risk level is labeled for each data set, and a risk level label is assigned to each data. These labels identify the risk status of customers in specific business activities. For example, a customer's history of not repaying on time in a lending activity will be labeled as a higher credit risk. Through this risk level labeling, a labeled training sample set is established, allowing the model to identify various risk signals.

[0050] Furthermore, after constructing the labeled training sample set, model training is performed based on these samples to obtain the final risk assessment model. During the model training process, machine learning algorithms such as logistic regression, decision trees, or neural networks are combined to learn the risk characteristics in the sample data. These algorithms find the best risk prediction model through iterative calculations and parameter optimization to maximize the accuracy of risk assessment. Model training will also continuously optimize parameters so that the model can make accurate predictions under different risk scenarios. For example, the decision tree algorithm can identify different risk characteristics of customers through a hierarchical structure, while the neural network can discover complex risk patterns through deep learning, thereby further improving the effect of risk assessment.

[0051] Furthermore, after the risk assessment model is trained and optimized, the model is called in real time during the business process. Based on the customer profile of the target customer, the relevant information is input into the risk assessment model, and the risk assessment results of the customer under various risk indicators are obtained. For example, by inputting the customer's financial data, transaction records, and historical risk records, the model can assess the customer's credit risk level, risk resistance in market fluctuations, and the possibility of operational risk. This real-time risk assessment not only improves the efficiency of financial services, but also adjusts the risk prediction results according to the latest customer data, so that the assessment is more in line with the customer's current business status.

[0052] Based on the solution of the present invention, by collecting and preprocessing historical data, model training is based on comprehensive and high-quality data sources, ensuring the reliability of the model. Based on the classification and labeling of different risk indicators, the model can identify various risk signals and improve the accuracy of risk prediction. The real-time call function of the model enables the system to dynamically evaluate the latest risk status of customers, thereby providing strong support for risk control and business decision-making of financial institutions. This technical solution not only improves the risk management capabilities of the system, but also significantly reduces the losses of financial institutions caused by inaccurate risk assessment, and ultimately achieves more accurate customer management and business recommendations.

[0053] Step S40: obtaining recommended services for target customers based on the customer portrait and the risk assessment result, and pushing the recommended services to the target client.

[0054] Specifically, based on collaborative filtering and / or content recommendation algorithms, recommended businesses for target customers are obtained based on the customer portrait and the risk assessment results; wherein the recommended business is any one or more of loan business, insurance business and leasing business.

[0055] In an embodiment of the present invention, a collaborative filtering algorithm is used to analyze the behavior patterns of similar customers to recommend services that have been selected by users with similar profiles to target customers. The collaborative filtering algorithm finds groups with similar characteristics to the target customers based on the basic information, financial status, and historical transaction preferences in the customer profiles. The business preferences of these groups will become the basis for recommending services to target customers. For example, if a customer's profile shows that he has a high demand for leasing services, and the risk assessment results show that he has the corresponding risk tolerance, leasing services will be recommended first. In addition, the collaborative filtering algorithm can continuously optimize the recommendation effect, because as more customer data is added, the algorithm will continuously update the similarity relationship between customers, thereby providing each customer with business recommendations that better meet their needs.

[0056] At the same time, the content recommendation algorithm will be combined to analyze the risk assessment results of the target customers and recommend more suitable business types for them. The content recommendation algorithm uses the customer's financial status, historical risk records and other characteristics to determine the customer's potential needs and suitability for loan, insurance and leasing businesses. For example, if the customer's risk assessment results show a low credit risk, a loan business will be recommended to him; if the customer's financial situation is stable, insurance business may be recommended to provide more comprehensive financial protection. This content-based recommendation method ensures that the recommended business is highly consistent with the customer's current situation, while reducing the probability of customers accepting higher-risk businesses and enhancing the security of the service.

[0057] Based on the solution of the present invention, by integrating collaborative filtering and content recommendation algorithms, it is possible to achieve precision and personalization of business recommendations, so that the recommendation results are more in line with the actual needs and risk tolerance of customers. This method not only improves the customer experience and enables them to obtain the services that best meet their conditions, but also significantly enhances the business matching rate and customer stickiness of financial institutions. In addition, the recommendation method based on customer portraits and risk assessment effectively reduces the customer churn rate caused by business mismatch, and ultimately achieves the goal of improving customer satisfaction and institutional business efficiency.

[0058] Figure 2 1 is a system structure diagram of a customer profile analysis system for financial services provided by an embodiment of the present invention. Figure 2As shown, an embodiment of the present invention provides a customer portrait analysis system for financial services, and the system includes: a collection unit, used to collect business-related information of a target customer in response to authorization information of the target customer; a processing unit, used to perform preprocessing on the business-related information, and construct a customer portrait of the corresponding target customer based on the preprocessed business-related information; a risk assessment unit, used to call a trained risk assessment model based on the customer portrait to obtain a risk assessment result of the corresponding target customer; a business recommendation unit, used to obtain recommended business for the target customer based on the customer portrait and the risk assessment result, and push the recommended business to the target client.

[0059] An embodiment of the present invention also provides a computer-readable storage medium, which stores instructions that, when executed on a computer, enable the computer to execute the above-mentioned customer profile analysis method for financial services.

[0060] Those skilled in the art will understand that all or part of the steps in the method for implementing the above-mentioned embodiments can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions for making a single-chip microcomputer, a chip or a processor (processor) perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0061] The optional embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the technical concept of the embodiments of the present invention, the technical scheme of the embodiments of the present invention can be subjected to a variety of simple modifications, and these simple modifications all belong to the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0062] In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed in the embodiments of the present invention.

Claims

1. A customer profile analysis method for financial services, characterized in that: The method comprises: In response to authorization information from a target customer, collect business-related information from the target customer; Preprocessing the business-related information, and building a customer profile of the corresponding target customer based on the preprocessed business-related information; Based on the customer profile, the trained risk assessment model is called to obtain the risk assessment result of the corresponding target customer; Based on the customer portrait and the risk assessment result, a recommended service for the target customer is obtained, and the recommended service is pushed to the target client.

2. The method according to claim 1, characterized in that The collecting of the target customer's business-related information in response to the target customer's authorization information includes: Receive and verify the authorization information of the target customer based on the SpringBoot framework; After verification, each microservice is called based on the SpringCloud distributed architecture to collect customer business-related information from multiple data sources in real time; The collected business-related information is temporarily stored in the Redis cache, and the cached data is written to the PostgreSQL database at predetermined intervals.

3. The method according to claim 1, characterized in that The preprocessing of the service related information includes: Execute the business related information cleaning based on SpringBoot framework; Convert the cleaned business-related information into a unified preset target format; Standardization processing is performed on the business-related information after format conversion to obtain pre-processed business-related information.

4. The method according to claim 1, characterized in that Build a customer profile of the target customer based on the pre-processed business-related information, including: Classify the preprocessed business-related information based on the predetermined portrait indicators to obtain a data set corresponding to each portrait indicator; Construct corresponding portrait indicators based on each data set; Based on various portrait indicators, a customer portrait of the target customer is obtained through combination.

5. The method according to claim 4, characterized in that The portrait indicators include: Any one or more of the following: basic customer information indicators, shareholder relationship indicators, foreign investment indicators, financial status indicators and historical risk assessment indicators.

6. The method according to claim 1, characterized in that The method further comprises: Perform risk assessment model training; The training rules of the risk assessment model are: Collecting historical business data and performing preprocessing on the historical business data; Classify the pre-processed historical business data based on each risk indicator, obtain the data set corresponding to each risk indicator, and perform risk level labeling on each data set; Training samples are constructed based on the labeled data sets, and model training is performed based on the training samples to obtain a risk assessment model.

7. The method according to claim 6, characterized in that Risk indicators include: Any one or more of credit risk, market risk and operational risk.

8. The method according to claim 1, characterized in that The obtaining a recommended service for a target customer based on the customer profile and the risk assessment result, and pushing the recommended service to a target client, includes: Based on collaborative filtering and / or content recommendation algorithms, recommended services for target customers are obtained based on the customer portrait and the risk assessment results; wherein, The recommended business is any one or more of loan business, insurance business and leasing business.

9. A customer profile analysis system for financial services, characterized in that: The system comprises: A collection unit, configured to collect business-related information of a target customer in response to authorization information of the target customer; A processing unit, configured to preprocess the business-related information and construct a customer profile of the corresponding target customer based on the preprocessed business-related information; A risk assessment unit, configured to call a trained risk assessment model based on the customer profile to obtain a risk assessment result corresponding to the target customer; A service recommendation unit is used to obtain recommended services for target customers based on the customer portrait and the risk assessment result, and push the recommended services to a target client.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the customer portrait analysis method for financial services described in any one of claims 1-8.

Images