Decision generation method and device based on big data, equipment and medium

Abstract

The invention relates to the technical field of big data and the field of financial science and technology, and discloses a decision generation method based on big data, and the method comprises the steps: collecting multi-dimensional data related to a target business, extracting key features, such as a behavior mode, a potential problem and resource stability, of a customer, constructing a customer analysis model, and carrying out the decision generation. Generating a time sequence data set in combination with the behavior trend data; a behavior data set is generated through scene analysis and quantitative analysis is carried out to obtain a behavior mode and a resource utilization trend of a client, and finally a strategy adapted to each business scene is generated. According to the method, the service strategy is finally generated by collecting the multi-dimensional data, extracting the customer characteristics, constructing the analysis model, analyzing the customer behavior trend, generating the time series data set and combining different service scenes to perform scene analysis and quantitative processing, so that the real-time analysis, scene decision and dynamic adjustment of the multi-dimensional data are realized; the flexibility and accuracy of the system in a complex service scene are improved, and the service decision-making efficiency and accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the fields of big data technology and financial technology, and in particular to a decision-making method, device, equipment and storage medium based on big data. Background Art

[0002] In the insurance industry, premium assessment is a crucial part of the underwriting process. Reasonable premium pricing not only helps insurance companies reduce losses and optimize business operations, but also provides customers with quotation plans that better meet their personal needs, thereby attracting more customers and improving market competitiveness. In current insurance business practices, there are mainly two common premium pricing methods: manual quotation and automatic quotation.

[0003] In the traditional manual quotation model, underwriters determine premiums based on their own experience and collected customer information. Although this method relies on underwriters' ability to judge and analyze customer information, it has the following significant shortcomings:

[0004] Reliance on personal experience: Manual quotations rely on the subjective judgment and experience of underwriters, so the accuracy and consistency of pricing results are difficult to guarantee.

[0005] Inefficiency: Underwriters need to review customer information one by one and process it manually, which makes it difficult to respond quickly to a large number of customer needs.

[0006] Limited data analysis: Manual methods have limited data processing capabilities and cannot fully utilize large amounts of historical data and market dynamics to optimize premium pricing.

[0007] With the rapid changes in the financial market, customer needs are becoming increasingly diverse and complex. Traditional premium pricing methods can no longer meet the needs of modern insurance business. Existing technologies lack sufficient flexibility and accuracy, and it is difficult to capture and analyze multi-dimensional customer behavior data and market changes in real time. Therefore, the existing premium assessment method can no longer effectively adapt to the current complex financial environment and customer needs. There is an urgent need for a pricing method that can dynamically adjust and flexibly respond to risk changes. Summary of the invention

[0008] The main purpose of the present invention is to provide a decision-making method, device, equipment and storage medium based on big data, aiming to solve the technical problem that the existing technology lacks the ability to analyze multidimensional data and customer behavior in real time, resulting in insufficient flexibility and accuracy of the system.

[0009] To achieve the above object, the present invention provides a decision making method based on big data, comprising:

[0010] Collect multi-dimensional data related to the target business, including customer personal information, historical behavior data, resource status and / or credit history;

[0011] Extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability;

[0012] Building a customer analysis model based on the customer characteristics;

[0013] Acquire key features of target customers, and use the customer analysis model to analyze behavioral trend data of the target customers based on the key features;

[0014] Segmenting the target customer's behavior trend data according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0015] Combine multiple business scenarios to perform scenario analysis on the time series data set to generate a corresponding behavior data set for each business scenario;

[0016] Quantitatively analyze the behavior data set corresponding to each business scenario to generate the behavior analysis results of target customers in each business scenario in each time period;

[0017] Based on the behavior analysis results of target customers in each business scenario in each time period, combined with the historical behavior data of target customers, target business strategies are generated for target customers.

[0018] Furthermore, to achieve the above-mentioned purpose, the present invention provides a decision-making device based on big data, comprising:

[0019] A data collection module, used to collect multi-dimensional data related to the target business, the multi-dimensional data including customer personal information, historical behavior data, resource status and / or credit record;

[0020] A feature extraction module, used to extract customer features that affect business decisions from the multidimensional data, wherein the customer features include customer behavior patterns, potential problem indicators and / or resource stability;

[0021] A customer analysis model building module, used to build a customer analysis model based on the customer characteristics;

[0022] A customer behavior analysis module, used to obtain key features of target customers and analyze the behavior trend data of the target customers based on the key features using the customer analysis model;

[0023] A time series processing module is used to process the behavior trend data of target customers in segments according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0024] A scenario analysis module, used to combine multiple business scenarios, perform scenario analysis on the time series data set, and generate a corresponding behavior data set under each business scenario;

[0025] The quantitative analysis module is used to perform quantitative analysis on the corresponding behavioral data set in each business scenario and generate the behavioral analysis results of the target customers in each business scenario in each time period;

[0026] The business strategy generation module is used to generate target business strategies for target customers based on the behavior analysis results of target customers in each business scenario in each time period and in combination with the historical behavior data of target customers.

[0027] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer device, which includes a memory, a processor, and a decision generation program based on big data stored in the memory and executable on the processor, wherein the decision generation program based on big data, when executed by the processor, implements the steps of the decision generation method based on big data as described above.

[0028] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, on which a decision generation program based on big data is stored, and when the decision generation program based on big data is executed by a processor, the steps of the decision generation method based on big data as described above are implemented.

[0029] Beneficial effects: The present invention relates to the fields of big data technology and financial technology, and discloses a decision-making method based on big data, which collects multidimensional data related to the target business, extracts key features such as customer behavior patterns, potential problems, and resource stability, builds a customer analysis model, and generates a time series data set in combination with behavior trend data; generates a behavior data set through scenario analysis and performs quantitative analysis to derive customer behavior patterns and resource utilization trends, and finally generates a strategy adapted to each business scenario. The present invention collects multidimensional data, extracts customer features, constructs an analysis model, analyzes customer behavior trends, generates a time series data set, performs scenario analysis and quantitative processing in combination with different business scenarios, and finally generates a business strategy, thereby achieving real-time analysis, scenario-based decision-making, and dynamic adjustment of multidimensional data, improving the flexibility and accuracy of the system in complex business scenarios, effectively reducing human intervention, and improving the efficiency and accuracy of business decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0031] Figure 1 A schematic diagram of an application environment of a decision making method based on big data in an embodiment of the present invention;

[0032] Figure 2 It is a flow chart of an embodiment of a decision making method based on big data of the present invention;

[0033] Figure 3 It is a functional module diagram of a preferred embodiment of a decision-making generation device based on big data of the present invention;

[0034] Figure 4 A schematic diagram of the structure of a computer device in one embodiment of the present invention;

[0035] Figure 5 FIG. 4 is another schematic diagram of the structure of a computer device in one embodiment of the present invention. DETAILED DESCRIPTION

[0036] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0037] The decision-making method based on big data provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In the application environment, the user end communicates with the server end through the network. The server end can collect multidimensional data related to the target business through the user end, extract key features such as customer behavior patterns, potential problems and resource stability, build a customer analysis model, and generate a time series data set in combination with behavior trend data; generate a behavior data set through scenario analysis and perform quantitative analysis to obtain customer behavior patterns and resource utilization trends, and finally generate strategies that adapt to various business scenarios. The present invention collects multidimensional data, extracts customer characteristics, builds an analysis model, analyzes customer behavior trends, generates a time series data set, performs scenario analysis and quantitative processing in combination with different business scenarios, and finally generates a business strategy, thereby realizing real-time analysis, scenario-based decision-making and dynamic adjustment of multidimensional data, improving the flexibility and accuracy of the system in complex business scenarios, effectively reducing human intervention, and improving business decision-making efficiency and accuracy. Among them, the user end can be but is not limited to various personal computers, laptops, smart phones, tablet computers and portable wearable devices. The server end can be implemented with an independent server or a server cluster composed of multiple servers. The present invention is described in detail below through specific embodiments.

[0038] See also Figure 2 , Figure 2This is a flow chart of an embodiment of a decision making method based on big data provided by the present invention. It should be noted that although a logical sequence is shown in the flow chart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0039] like Figure 2 As shown, the decision-making method based on big data proposed in the present invention includes the following steps:

[0040] S10, collecting multi-dimensional data related to the target business, wherein the multi-dimensional data includes customer personal information, historical behavior data, resource status and / or credit record;

[0041] In this embodiment, customer personal information refers to the basic identity information of the customer, including but not limited to name, ID number, contact information, address, etc. In the financial field, customer personal information is one of the basic data for determining risk assessment, and can be used to analyze customer stability, credit assessment, etc. By connecting to the customer data storage system, calling the API interface or database query method, batch acquisition and update of this personal information from the customer database. At the same time, during the data collection process, the system needs to perform identity authentication to ensure that the data source is legal and accurate.

[0042] Historical behavior data includes the customer's past operations and behavior records, such as transaction records, login history, and the frequency of using certain services. This type of data can be used to analyze customer behavior patterns and predict future behavior trends. Historical behavior data is usually stored in behavior logs. The system connects to the customer behavior log data source through log analysis tools or big data analysis platforms to regularly collect and cleanse this data. During the data cleaning process, the system should eliminate invalid data to ensure the validity of the data.

[0043] Resource status refers to the resources currently available to the customer, including funds, assets, liquidity status, etc. In financial data, resource status can reflect the financial health of the customer and directly affect the risk assessment of business decisions. By connecting to a third-party financial institution or corporate financial management system, the system can obtain data such as the customer's capital flow and balance sheet. This data can be regularly updated and collected through the API interface or financial data service platform.

[0044] Credit records are the credit history of customers, including their credit scores, overdue records, repayment records, etc. These data are important bases for financial institutions to conduct risk control and credit assessment. Credit records can be obtained by connecting to third-party credit platforms (such as credit reporting agencies). Through regular credit data synchronization, the system can call the credit query interface to automatically update the customer's credit status. Data security and privacy protection need to be ensured, and the acquisition of credit records must comply with relevant laws and regulations.

[0045] Example: When a financial institution approves a customer's loan, it collects the customer's personal information (such as age, occupation), historical behavior data (such as previous loan repayments), resource status (such as current income and assets), and credit record (such as credit score, number of overdue payments) through the system, comprehensively assesses the customer's repayment ability and risk from multiple dimensions, and provides intelligent decision-making suggestions based on the system analysis results. This technical means based on multi-dimensional data analysis can significantly reduce the credit risk of financial institutions and improve the efficiency of credit approval.

[0046] By collecting multi-dimensional data, the system can conduct a comprehensive analysis of customers from multiple dimensions such as customer personal information, historical behavior data, resource status and credit records, and accurately predict customer behavior patterns and potential risks. This not only improves the automation of data collection and reduces manual operations, but also improves the real-time and integrity of data, thereby providing more reliable data support for subsequent business decisions and improving the efficiency and accuracy of business decisions.

[0047] S20, extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability;

[0048] In this embodiment, the behavior pattern refers to the customer's habitual behavior and frequency in past operations. These patterns may include the customer's consumption behavior, transaction frequency, interaction mode, etc., which can help the system predict the customer's future behavior trends, thereby affecting business decisions. The extraction of behavioral patterns helps the system identify the customer's regular operating behaviors and preferences. By analyzing historical behavioral data, the system uses machine learning models or data mining techniques to model and classify customers' daily interaction records (such as transaction frequency, consumption amount, and visit time). The system can perform label classification according to different types of behavioral patterns to generate the "behavioral pattern" part of the customer portrait.

[0049] Potential problem indicators refer to characteristics that may indicate that a customer has potential risks. These include, but are not limited to, a customer's reduced ability to perform, abnormal behavior, and historical default records. Potential problem indicators can help the system identify risky customers in a timely manner and make corresponding adjustments in business decisions, such as adjusting risk control strategies or restricting specific business operations. The system identifies possible risk characteristics by analyzing the customer's credit record, resource status, and historical behavior data. For example, if a customer has frequently changed contact information or has suddenly reduced transaction behavior recently, the system will mark these behaviors as "potential risks." By comparing past normal behavior patterns, the system can automatically identify these abnormal behaviors and generate a list of "potential problem indicators."

[0050] Resource stability refers to whether the client's available resources (such as funds, assets, liquidity) are stable. Resource stability directly affects the client's ability to perform and is therefore a key factor in business decision-making. Clients with high resource stability generally mean lower risk, while clients with unstable resources may have higher credit risk. By analyzing the client's resource status (such as bank account balance, assets and liabilities, investment status, etc.), the system can regularly evaluate the client's resource fluctuations. The system can generate a "resource stability" score or classification label based on the client's capital flow trend, total asset changes, etc. The scoring or labeling results can be used in further business decision-making processes.

[0051] Example: A customer has maintained a stable transaction frequency and capital flow in the past 6 months, and the system marks his behavior pattern as "stable". However, in the past month, the number of overdue credit card payments of this customer has suddenly increased, and the balance of the capital account has fluctuated greatly. By extracting its "potential problem indicators" and "resource stability" characteristics, the system recommends that the bank adjust the credit limit of this customer and issue a risk warning. This technical means based on multidimensional data analysis enables banks to make business decisions more timely and accurately, and reduce credit risks.

[0052] By extracting customer behavior patterns, potential problem indicators, and resource stability from multidimensional data, the system can comprehensively analyze the key characteristics of customers and accurately assess customer risks and behaviors. It can improve the automation of data analysis, reduce manual intervention, enhance the accuracy and flexibility of business decisions, and ensure that enterprises can promptly identify potential risks and adjust business strategies, thereby reducing possible losses.

[0053] S30, building a customer analysis model based on the customer characteristics;

[0054] In this embodiment, before building a customer analysis model, the extracted customer feature data must first be cleaned and preprocessed. Data cleaning mainly includes removing duplicate data, processing missing values ​​and outliers, etc., and data preprocessing includes operations such as data formatting, standardization and normalization. This process ensures that the data input to the model is accurate, neat and suitable for subsequent modeling. The system can use common data preprocessing tools or libraries (such as Pandas, Scikit-learn, etc.) to standardize the original customer feature data. For example, the customer's income, transaction amount, etc. are normalized so that data with different characteristics can be processed on the same scale, thereby improving the training effect of the model.

[0055] Customer analysis models can be constructed using a variety of machine learning and artificial intelligence models. The specific model type should be selected based on the complexity of customer characteristics and business needs. Common basic model types include decision trees, support vector machines (SVMs), neural networks, etc. Different models are suitable for different scenarios. For example, decision trees are suitable for simple decision-making problems, while neural networks are suitable for dealing with complex nonlinear problems. The system automatically selects suitable basic models based on the structure and quantity of different customer characteristics. For example, when the customer's historical behavior pattern data is more complex, a neural network model based on deep learning can be selected; for simple classification problems, a decision tree or random forest model can be used. Model selection can be optimized through technical means such as cross-validation and model evaluation.

[0056] In order to ensure the accuracy and generalization ability of the model, customer feature data needs to be divided into training set, validation set and test set. The training set is used for model training, the validation set is used for model parameter adjustment and evaluation, and the test set is used for the final performance evaluation of the model. The system divides customer feature data according to a certain ratio (such as 70% training set, 20% validation set, 10% test set). The division can be randomly assigned to ensure the diversity and representativeness of the data set. The divided data will be used for model training and evaluation at different stages.

[0057] Model training is to input customer feature data into the selected basic model and learn the relationship between customer features and targets through the model training process. Model optimization includes adjusting the parameters of the model (such as learning rate, regularization parameters, etc.) to improve the accuracy and generalization ability of the model. The system inputs the training set into the basic model and uses the machine learning algorithm to train the model. During the training process, the system iteratively updates the model parameters to minimize the loss function (such as mean square error or cross entropy). After the training is completed, the system will use the validation set to evaluate the performance of the model and adjust the hyperparameters based on the validation results to obtain the optimal model.

[0058] The optimized model needs to undergo a final performance evaluation to ensure that the model can achieve good results in practical applications. Evaluation indicators include accuracy, precision, recall, F1-score, etc. After the model evaluation is completed, the trained model can be deployed to the business system for real-time customer analysis. The system uses the test set to evaluate the performance of the model, calculate various evaluation indicators, and compare them with the evaluation results of the validation set to ensure the generalization ability of the model. After the evaluation is completed, the system will deploy the model to the production environment for real-time analysis of customer data and generation of business decisions.

[0059] Example: A bank wants to conduct credit assessment for new loan applicants. The system builds a customer analysis model by collecting credit records, historical repayment data, income changes and other features of multiple customers. By training data from multiple customers, the model can identify which customers have good repayment ability and which customers have potential risks. For new loan applicants, the system inputs their key features into the trained model. The model determines the customer's future repayment ability based on historical data, predicts the possibility of loan default, and generates a business strategy for the bank that adapts to the current risk situation.

[0060] By building a customer analysis model based on customer characteristics, the system can accurately capture customer behavior patterns, potential problems, resource stability and other characteristics, and use machine learning or artificial intelligence technology to conduct intelligent analysis and prediction of customers. The model can be dynamically adjusted and optimized to enhance the flexibility and accuracy of the system and help enterprises make accurate business decisions in complex business scenarios.

[0061] S40, acquiring key features of target customers, and using the customer analysis model to analyze the behavior trend data of the target customers based on the key features;

[0062] In this embodiment, the key features of the target customers refer to important customer features that can affect business decisions. These features usually include customer behavior patterns, potential problem indicators, resource stability, etc., and are representative data extracted from the multidimensional data of the target customers. The purpose of extracting these key features is to better capture the individual behavioral characteristics of customers, so as to make more accurate analysis and predictions. The system automatically extracts the relevant data of the target customers through the constructed customer analysis model, and filters out the features that have a significant impact on business decisions. This can be achieved through feature selection algorithms (such as recursive feature elimination or feature importance based on tree models). Key features can include historical transaction frequencies, risk points in credit records, resource changes, etc.

[0063] Customer behavior trend data is based on the key features of target customers, predicting their possible behavioral changes in the future. Behavioral trend analysis aims to predict customers' future behavior patterns based on their historical behaviors and key features, helping the business make predictive decisions. These behavioral trends may include changes in consumption patterns, fluctuations in resource utilization, and increases in potential risks. By inputting the key features of target customers into a trained customer analysis model, the system can make predictions based on these features. The customer analysis model may be trained using a supervised learning algorithm (such as a support vector machine, neural network, or random forest, etc.), which analyzes the customer's behavioral trends in the future based on the target customer's historical data and key features, and outputs trend analysis results. For example, the model can predict the customer's future consumption behavior frequency or resource utilization changes.

[0064] Behavior trend data is the prediction result of target customers' behavior in a preset time period in the future. This data will show the changes in customers' behavior at different time points, helping enterprises adjust their business strategies according to customers' future trends. Based on the output of the customer analysis model, the system generates behavior trend data, covering prediction results for multiple time periods. This data will be recorded in the time series database and used for subsequent scenario analysis and business strategy formulation.

[0065] Example description: Insurance companies automatically obtain key features of auto insurance customers such as historical driving records, accident frequency, and vehicle maintenance through the system. Based on these features, the customer analysis model predicts the customer's possible future driving behavior trends. For example, a customer's historical records show that he has a high driving frequency and has had many minor accidents in the past year. The system predicts that he is more likely to have an accident in the next year. Insurance companies can adjust premiums based on this behavioral trend, or recommend safe driving training courses to customers, thereby reducing the risk of future claims.

[0066] By acquiring the key characteristics of target customers and using customer analysis models to analyze behavioral trends, the system can accurately predict customers' future behavioral patterns and resource changes. This predictive analysis enables companies to adjust their business strategies based on customers' future trends, thereby better responding to market changes and customer needs, improving business flexibility and accuracy, and reducing potential risks.

[0067] S50, segmenting the behavior trend data of the target customers according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0068] In this embodiment, the segmentation of the time period refers to dividing the target customer's behavior trend data according to a set time period, such as by day, week, month or year. The purpose of the segmentation is to divide the customer's behavior trend data into multiple time segments, so that a more detailed analysis of the changes in the customer's behavior at different time points can be performed. The setting of the time period can be flexibly configured according to business needs or the characteristics of the data. The system pre-sets the time period and divides the customer's behavior data into different time periods according to the specified period based on the business needs of the target customer. The system can automatically extract the data in each time period and archive the data to ensure the independence and continuity of the data. During the data segmentation process, the system should take into account the integrity of the data to ensure that the data in each time period is not missing or omitted.

[0069] A time series dataset refers to a behavioral trend dataset arranged in chronological order. A time series dataset contains behavioral trend data over multiple time periods and can show the behavioral changes of target customers over different periods. The purpose of generating a time series dataset is to be able to analyze the changing trends of customer behavior over different periods, thereby providing support for subsequent scenario analysis and decision-making. The system arranges the behavioral data in each time period in chronological order to generate a complete time series dataset. This dataset covers the behavioral changes of customers over multiple time periods, ensuring the consistency of data on the timeline. The system can mark the position of each data segment with a timestamp or period identifier, and store the time series data in the corresponding database for subsequent analysis and processing.

[0070] In a time series data set, there may be missing or abnormal data points. The system needs to process this data to ensure the integrity and accuracy of the time series data. Missing data is usually completed by interpolation or regression methods, and abnormal data can be corrected by filtering or smoothing. In the process of generating a time series data set, the system first checks the integrity of the data. For missing data points, the system can use interpolation methods (such as linear interpolation, spline interpolation, etc.) to complete the data. For abnormal data points, the system can perform smoothing based on context or historical data to ensure the continuity of the data.

[0071] Example: An insurance company analyzes the driving behavior of auto insurance customers. By collecting the customer's historical driving records, the system processes the data by month and generates a time series data set. The data set contains the customer's monthly driving mileage, number of traffic violations, vehicle maintenance status, etc. The system analyzes the time series data and finds that the customer's number of violations has increased significantly in the past few months, and the frequency of vehicle maintenance has decreased. Based on this trend analysis, the insurance company predicts that the customer's risk of accidents in the future will increase, and decides to adjust its premium or recommend that the customer participate in driving safety training.

[0072] By segmenting the target customer's behavior trend data according to the time period and generating a time series data set, the system can dynamically track the changing trend of customer behavior. The time series data set provides the customer's behavior performance in multiple time periods, allowing enterprises to predict future behavior patterns and formulate more accurate business strategies based on the customer's historical trends. It improves the precision of data analysis and the scientific nature of decision-making, ensuring flexible response and optimized management of the business.

[0073] S60, combining multiple business scenarios, performing scenario analysis on the time series data set to generate a behavior data set corresponding to each business scenario;

[0074] In this embodiment, a business scenario is an operation or behavior scenario under specific conditions or environments defined according to target business needs. These scenarios may include routine operations, emergency handling, resource fluctuations, etc. By defining multiple business scenarios, the system can analyze the behavioral data of target customers for different scenarios, thereby more accurately predicting the performance of customers in different environments. The system first predefines multiple business scenarios based on business needs and actual application environments. Each business scenario is assigned a unique identifier to distinguish data analysis in different scenarios. For example, in the insurance field, the scenarios may be "customers use insurance products normally", "customers claim for sudden accidents", "customer behavior when the market environment fluctuates drastically", etc.

[0075] Scenario analysis refers to analyzing the behavior patterns of target customers in different business scenarios. Scenario analysis can help the system identify changes in customer behavior in different situations, so as to make more targeted assessments of customer behavior. This analysis can reveal the behavioral characteristics of customers under specific conditions and help companies make personalized business decisions. The system analyzes each business scenario separately based on the behavioral trend data in the time series data set. For example, in the "sudden accident claims" scenario, the system can analyze the customer's behavioral patterns such as accident frequency, claims needs and processing time. The analysis results of each scenario are output separately and used to generate the behavioral data set for the scenario.

[0076] A behavioral dataset is a complete dataset generated through scenario analysis that reflects the behavioral characteristics of customers in a specific business scenario. Each behavioral dataset reflects the behavioral patterns and characteristics of customers in a certain scenario and can provide support for subsequent business decisions. Based on the results of scenario analysis, the system generates behavioral datasets for multiple scenarios. Each dataset corresponds to a business scenario and covers information such as the customer's behavioral patterns, resource utilization, and risk assessment in that scenario. For example, in the "market fluctuation" scenario, the behavioral dataset generated by the system may contain data such as changes in customer purchasing behavior and risk tolerance.

[0077] Example: An insurance company conducts routine operation scenario analysis on the behavior of auto insurance customers. The system analyzes customers' daily driving habits, maintenance records, and insurance product usage through time series data sets. The system finds that a customer's driving habits in the past three months show high-risk characteristics, such as frequent speeding and increased accidents. Based on the behavioral data sets of routine operation scenarios, the insurance company can recommend customers to receive driving safety training and adjust the premium appropriately.

[0078] By combining multiple business scenarios to analyze time series data sets, the system can comprehensively evaluate customer behavior in different scenarios. The behavioral data sets generated in each scenario provide precise support for enterprises to make business decisions in different situations. This scenario-based analysis enables the system to not only respond to regular business needs, but also provide more flexible business strategies in emergencies and market fluctuations, improving the accuracy and response speed of customer service.

[0079] S70, performing quantitative analysis on the behavior data set corresponding to each business scenario, and generating behavior analysis results of target customers in each business scenario in each time period;

[0080] In this embodiment, quantitative analysis refers to converting qualitative behavioral data into quantifiable values ​​for statistical analysis. The system generates behavioral data sets for target customers in each business scenario, which usually contain multi-dimensional behavioral characteristics (such as resource utilization, risk indicators, behavioral patterns, etc.). Through quantitative analysis, the system can perform statistics, scoring, modeling, and other processing on these behavioral data in order to more accurately evaluate the performance of customers in specific scenarios. The system can use statistical methods (such as mean, variance, distribution analysis, etc.) or machine learning algorithms (such as cluster analysis, regression analysis, etc.) to perform quantitative analysis on behavioral data sets. For example, in the auto insurance business scenario, the system scores the customer's driving behavior (such as the number of speeding, accident frequency), converts qualitative behavior into quantitative indicators, and generates behavioral scores or trend change data. This process ensures that the behavioral data of each customer is comparable and evaluable in different scenarios.

[0081] The behavior analysis results are obtained by comprehensively evaluating the quantified behavior data, reflecting the customer's behavior performance in a specific time period and business scenario. The behavior analysis results can include the customer's behavior pattern, risk level, resource utilization efficiency, etc. These results provide an important basis for subsequent business decisions. The system generates the behavior analysis results for each time period based on the quantitative analysis results in each business scenario. For example, in the auto insurance business scenario, the system analyzes the customer's driving behavior data in the past three months and derives the accident risk trend of the customer in a specific time period. The system classifies or rates customers based on the quantitative results in order to provide insurance companies with targeted pricing or service recommendations.

[0082] The results of behavioral analysis conducted in multiple time periods need to be comprehensively analyzed in combination with the time dimension. This means that the system must not only quantify the behavior in a single time period, but also conduct cross-period analysis of data from multiple time periods to discover changing trends or long-term performance of customer behavior. The system vertically integrates the behavioral analysis results generated in each time period in order to analyze the changing trends of customer behavior. For example, the system can perform a time series analysis of a customer's driving behavior over the past year to identify changes in their driving habits or trends in increased risk. These cross-period behavioral analysis results can serve as a basis for long-term decision-making.

[0083] Example: An insurance company wants to conduct quantitative analysis on the driving behavior of auto insurance customers. The system generates a driving behavior score by quantifying the customer's speeding times, emergency braking times, and accident frequency. The system analyzes the customer's driving data over the past three months and finds that the customer's speeding behavior occurs frequently every weekend, increasing the risk of accidents. Based on the quantitative analysis results, the insurance company provides the customer with driving behavior improvement suggestions and adjusts his premium when renewing the policy.

[0084] By quantitatively analyzing the behavioral data set in each business scenario, the system can convert qualitative customer behavior data into actionable numerical indicators and generate customer behavior analysis results in each time period. These quantitative analysis results provide companies with more accurate customer behavior assessments and help them develop more effective business strategies in multiple business scenarios. Based on these analysis results, companies can adjust customers' insurance products, risk assessments, and pricing strategies, thereby improving business flexibility and customer satisfaction.

[0085] S80, based on the behavior analysis results of the target customers in each business scenario in each time period and in combination with the historical behavior data of the target customers, a target business strategy is generated for the target customers.

[0086] In this embodiment, the system comprehensively evaluates the performance of customers in specific situations based on the behavior analysis results of target customers in different time periods and multiple business scenarios. These behavior analysis results include customer behavior patterns, resource utilization, risk assessment, etc., which can provide data support for subsequent business strategies. The system conducts a global evaluation based on the behavior analysis results of customers in different business scenarios in each time period. For example, in the auto insurance scenario, the system analyzes the driving behavior data of customers in each month, and derives the risk level of the customer based on the frequency of accidents and driving habits. This comprehensive behavior analysis can reflect the full picture of the customer and provide a reference for the generation of subsequent business strategies.

[0087] The system not only uses the behavior analysis results of the customer in the current time period, but also combines it with its historical behavior data for comparative analysis to evaluate the changing trend of customer behavior. Historical behavior data can be a customer's long-term behavior record, including past behavior performance in similar scenarios. Through this comparison, the system can identify long-term changes in customer behavior and thus generate business strategies more accurately. The system compares the behavior analysis results of each time period with the customer's historical data. For example, in the field of health insurance, the system can compare the customer's recent health status with his long-term health record to determine the improvement or deterioration trend of the customer's health status. Through this comparative analysis, the system can better predict the customer's future behavior and provide a more accurate basis for the formulation of business strategies.

[0088] Based on the comparison of comprehensive behavioral analysis results and historical behavioral data, the system generates personalized business strategies suitable for target customers. Business strategies can be suggestions for optimizing customer resource allocation, providing personalized insurance products, or adjusting customers' risk levels, premiums, etc. The target business strategy is to help companies flexibly adjust business content according to changes in customer behavior. After a comprehensive analysis of the target customer's behavioral data, the system generates specific business strategies based on the evaluation results. For example, in a life insurance scenario, the system can recommend flexible payment plans for customers based on their financial changes, or adjust the insurance amount based on their risk status. The generated strategies directly affect subsequent business decisions and help companies optimize services and manage customer relationships.

[0089] Example: A car insurance company systematically analyzes a customer's driving behavior and accident records, and combined with his historical driving habits, finds that the customer's accident frequency has increased in the past three months. The business strategies generated by the system include increasing premiums, recommending driving safety training courses, or providing additional safety equipment. These strategies help insurance companies reduce future claims risks while providing customers with targeted safety improvement suggestions.

[0090] By combining the behavioral analysis results in each business scenario in each time period, the system can comprehensively evaluate the customer's behavior patterns and risk changes. Combined with historical behavioral data for comparative analysis, the system can not only identify the customer's current status, but also predict its future behavioral trends, thereby generating personalized business strategies for customers. This strategy generated based on multi-dimensional data analysis helps companies flexibly adjust business strategies in different scenarios, improve the accuracy and response speed of customer service, and effectively reduce the company's operational risks.

[0091] The present invention relates to the field of big data technology and financial technology, and discloses a decision-making method based on big data, which collects multidimensional data related to the target business, extracts key features such as customer behavior patterns, potential problems and resource stability, builds a customer analysis model, and generates a time series data set in combination with behavior trend data; generates a behavior data set through scenario analysis and performs quantitative analysis to derive customer behavior patterns and resource utilization trends, and finally generates a strategy adapted to each business scenario. The present invention collects multidimensional data, extracts customer features, constructs an analysis model, analyzes customer behavior trends, generates a time series data set, performs scenario analysis and quantitative processing in combination with different business scenarios, and finally generates a business strategy, thereby achieving real-time analysis, scenario-based decision-making and dynamic adjustment of multidimensional data, improving the flexibility and accuracy of the system in complex business scenarios, and improving the efficiency and accuracy of business decision-making.

[0092] In one embodiment, in the above S10, collecting multi-dimensional data related to the target business includes:

[0093] S101, identifying a multidimensional data source related to a target business, wherein the multidimensional data source includes a customer database, a credit inquiry platform and / or a public resource data platform;

[0094] S102, configuring corresponding data collection interfaces for different data sources, wherein the data collection interfaces are used to obtain customer personal information, historical behavior data, resource status and / or credit records from the data sources;

[0095] S103, performing unified formatting processing on the collected data, converting the data from different data sources into a standardized data structure, wherein the unified formatting processing includes unifying the time format and converting the numerical unit.

[0096] In this embodiment, the multidimensional data source refers to a data source related to the target business and obtained from multiple channels, including a customer database, a credit inquiry platform, and a public resource data platform. Different data sources provide customer information of different dimensions, which helps to form a global understanding of the customer. For example, the customer database can provide the basic information of the customer, the credit inquiry platform can provide the credit record, and the public resource data platform can provide the financial or resource status of the customer. The system first identifies the data source related to the target business through configuration and determines which external or internal platforms need to obtain data from. The system connects the customer database, third-party credit platform (such as a credit rating company) and public resource data platform (such as a government website, market research platform), etc., to ensure that the multidimensional information of the target customer can be fully collected.

[0097] The data collection interface is an interface module used to obtain customer information from various data sources. Each data source may have different formats, protocols, and access methods, so it is necessary to configure the corresponding interface for each data source to ensure that data can be smoothly and accurately collected from different platforms into the system. For example, the customer database can use API calls, while the public resource data platform may collect data by regularly importing data tables. The system designs and develops corresponding data collection interfaces for each data source. For the customer database, the system may use a database connection driver to directly access the data; for the credit inquiry platform, the system may call through the API interface; for the public resource data platform, the system may use a data crawler or a regular update function for data collection. Each interface module is designed according to the characteristics of the data source to ensure the real-time and integrity of data collection.

[0098] Different data sources usually provide data in different formats, which may include different time formats, numerical units, etc. In order to ensure that the system can process and analyze multidimensional data from different data sources, unified formatting is required to standardize all data into a unified structure. The formatting process includes unified time format, numerical unit conversion, etc. to ensure the consistency and comparability of data. After data collection, the system preprocesses all data. First, the system checks the time format of the data and converts it to a unified standard format (such as ISO 8601 format). Then, the system performs standardized conversion of numerical units to ensure that all data is expressed in consistent units. For data in different units, the system can convert them through predefined conversion rules, such as converting US dollars to local currency units. The formatted data can be seamlessly connected in subsequent analysis.

[0099] This embodiment collects multidimensional data related to the target business from multiple data sources and formats the data in a unified manner, so that the system can provide comprehensive and standardized data support. The integration and standardization of multidimensional data enables the system to efficiently process data across multiple data sources, improving the availability and consistency of data. This ensures that data analysis and decision-making in different business scenarios are more accurate, helping enterprises to formulate personalized product and service strategies.

[0100] In one embodiment, the above S20 includes:

[0101] S201, analyzing the historical behavior data of the customer to determine the behavior pattern of the customer, wherein the behavior pattern includes the customer's usage habits, interaction frequency, and access behavior;

[0102] S202, analyzing the credit record of the customer to determine potential problem indicators of the customer, wherein the potential problem indicators include the possibility of insufficient performance ability and abnormal behavior of the customer;

[0103] S203, analyzing the client's resource status and determining the client's resource stability, wherein the resource status includes the client's resource source, expenditure status, and available resources;

[0104] S204, integrating the customer's behavior pattern, potential problem indicators and resource stability to generate comprehensive customer characteristics for business decision-making.

[0105] In this embodiment, historical behavior data refers to the customer's interaction, usage records and other behavioral information within a specific time period. By analyzing this data, the system can identify the customer's behavior patterns, such as usage habits, interaction frequency and access behavior. These behavior patterns help companies understand customer preferences and behavioral patterns, thereby providing a basis for business decisions. The system extracts specific indicators from the customer's historical behavior data, such as daily access frequency, interaction behavior (such as purchase, inquiry) and usage habits (such as product access time period, usage frequency). The system can use data analysis tools and models (such as cluster analysis, association rule mining, etc.) to identify customer behavior patterns. For example, in the insurance industry, analyze the customer's claims records, policy management behaviors, etc. to determine the customer's insurance usage habits and frequency.

[0106] Credit records reflect a customer's performance history and credit risk. The system analyzes a customer's credit record to identify potential risks and problems, and to determine the possibility of a customer's inability to perform and abnormal behavior. Potential problem indicators can help companies warn of a customer's potential default risk or abnormal behavior in advance. The system obtains a customer's credit history (such as credit score, arrears record, etc.) from a third-party credit inquiry platform or internal credit data, and identifies a customer's credit risk by analyzing changes in the customer's credit score, loan performance, etc. For example, in the insurance industry, the system can evaluate a customer's ability to perform contracts in the future and identify potential default risks by analyzing the customer's loan performance history and debt situation.

[0107] Resource status refers to the stability of the client in terms of economy, finance, materials, etc. By analyzing the client's resource sources, expenditures, and available resources, the system can assess the client's resource stability. This feature directly affects the client's decision-making ability and risk tolerance in future business. The system assesses the stability of its resources by analyzing the client's resource data, such as income sources, expenditure records, available resources, etc. For example, the system can analyze the client's monthly income, expenditure stability, and assets to calculate the client's resource health and risk tolerance. For example, in the insurance field, the client's income and expenditure status are analyzed to assess whether the client has the ability to continue to pay insurance premiums.

[0108] The purpose of integrating customers' behavior patterns, potential problem indicators, and resource stability is to form a comprehensive customer profile to help companies make more accurate decisions. By comprehensively considering these different dimensions of characteristics, companies can better assess customers' business value, risk status, and the possibility of future business cooperation. The system combines three types of characteristics: customer behavior patterns, potential problem indicators, and resource stability to form comprehensive customer characteristics. These comprehensive characteristics can be quantified through a scoring model to facilitate subsequent business decisions. For example, the system integrates a customer's insurance product usage behavior, credit risk, and income stability to generate a comprehensive score that can be used to tailor products for customers or adjust premium structures.

[0109] This embodiment can provide more comprehensive support for business decision-making by extracting customer behavior patterns, potential problem indicators and resource stability from multidimensional data. The system can identify customer behavior patterns, potential risks and resource health, helping enterprises to formulate more accurate business strategies. It can effectively improve the risk management capabilities of enterprises and enhance customer satisfaction and business growth potential.

[0110] In one embodiment, the above S30 includes:

[0111] S301, determining a basic model according to the complexity of the customer characteristics and the target business requirements, the basic model including a decision tree, a support vector machine or a neural network;

[0112] S302, dividing the customer feature data into a training set and a validation set;

[0113] S303, inputting the training set of customer feature data into the basic model, performing preliminary model training, and optimizing the parameters of the basic model through iteration;

[0114] S304, using the validation set to analyze the trained model, determine the performance indicator results of the model on the validation set, and adjust the hyperparameters of the model according to the performance indicator results to generate a trained customer analysis model.

[0115] In this embodiment, the basic model is a statistical or machine learning model selected based on the diversity and complexity of customer characteristics and the business needs of the enterprise. The choice of the basic model directly determines the effect of subsequent model training and the accuracy of business applications. Common basic models include decision trees, support vector machines (SVMs) and neural networks. The system first selects a suitable basic model based on the scale of customer feature data, data dimensions and business objectives. For example, decision trees are suitable for scenarios where the data structure is relatively simple and the decision path can be expressed in a tree structure; support vector machines are suitable for classification tasks of small-scale high-dimensional data; neural networks are suitable for scenarios that process large-scale, complex data, such as predicting customer behavior trends or long-term customer churn rates.

[0116] When building a machine learning model, the data set usually needs to be divided into a training set and a validation set. The training set is used to train the model, and the validation set is used to evaluate the performance of the model to ensure that the model does not overfit the training data and can perform well on unseen data. The division of the data set usually follows a certain ratio, such as 80% of the data as a training set and 20% of the data as a validation set. The system randomly divides the collected customer feature data, using most of the data as a training set and the rest as a validation set. The system ensures that the training set covers as many customer feature situations as possible, and the validation set is used for a preliminary evaluation of the model performance. In the auto insurance scenario, the system may divide the customer's driving data, credit score, etc. into time periods to ensure that the time and data distribution between the training and validation sets are representative.

[0117] Model training is the process of learning and adjusting model parameters based on the structure of the model using the data in the training set. The system uses customer feature data to perform preliminary training on the basic model and adjusts the internal parameters of the model (such as the branch structure of the decision tree, the boundaries of the support vector machine, or the weights of the neural network) to improve the model's ability to recognize feature patterns. The system inputs the training set data into the selected basic model. For example, in the decision tree model, the system constructs a decision tree by splitting customer features; in the neural network, the system adjusts the weights of the network through multiple rounds of iterations. Through this process, the system continuously optimizes the parameters of the model so that it can better classify or predict customer features.

[0118] The purpose of using a validation set is to evaluate the performance of a trained model on unseen data. The validation set can help evaluate the generalization ability of the model and provide feedback for model optimization. The system can detect whether the model has overfitting problems through evaluation on the validation set. The system inputs the validation set into the trained model and calculates the model's performance indicators, such as accuracy, recall, F1 score, etc. In the auto insurance scenario, the system may measure the performance of the model by verifying the accuracy of the customer's driving behavior prediction. Based on the analysis results on the validation set, the system evaluates whether the model's performance meets business requirements.

[0119] Hyperparameter adjustment is to optimize the overall performance of the model. Hyperparameters include the model's structure, learning rate, regularization parameters, etc. The system adjusts the model's hyperparameters based on the performance results on the validation set and ultimately generates an optimized model. This process is usually completed through multiple iterations and verifications. The system adjusts the model's hyperparameters based on the analysis results of the validation set. For example, for a neural network model, the system can adjust the number of hidden layers or the learning rate; for a support vector machine, the system can adjust the kernel function or the regularization coefficient. The system optimizes the hyperparameters through multiple iterations to generate a customer analysis model with the best performance. After verification, the final customer analysis model can be used to predict customer behavior, risk assessment, and other tasks.

[0120] In this embodiment, by building a customer analysis model based on customer characteristics, the system can better predict customer behavior and assess business risks. Through model training of multi-dimensional characteristics and performance evaluation of the validation set, the system can ensure that the model has good generalization ability and can generate personalized business strategies based on the characteristics of different customers.

[0121] In one embodiment, the above S40 includes:

[0122] S401, collecting the behavior data and resource status of target customers through the real-time data monitoring module;

[0123] S402, generating an interaction relationship diagram between the target customer and external related entities based on the behavior data and resource status of the target customer;

[0124] S403, analyzing the interaction behavior between the target customer and the external related entity based on the interaction relationship diagram, and determining the key features that affect the target customer's behavior;

[0125] S404, inputting the key features into a customer analysis model to analyze the behavior trend data of the target customer, wherein the behavior trend data reflects the behavior analysis results of the target customer in multiple time periods.

[0126] In this embodiment, the real-time data monitoring module is responsible for collecting the target customer's behavior data in a specific business and its resource status. These data may include the customer's usage behavior (such as access frequency, interactive operations, etc.) and resource status (such as account balance, credit status, asset information, etc.). Through real-time monitoring, the system can timely capture changes in customer behavior and analyze its impact on the business. The system deploys a real-time data monitoring module to regularly collect customers' real-time behavior data and resource status. For example, an insurance company can monitor a customer's policy operation record, the frequency of initiation of claims requests, account balances, etc. At the same time, by connecting with the data of a third-party resource platform, the system can also obtain the customer's credit record or other resource-related information. These data serve as the basis for subsequent behavioral analysis to ensure that the analysis is timely and accurate.

[0127] The interaction relationship diagram is a visual diagram used to represent the interaction relationship between target customers and external related entities (such as partners, service providers, suppliers, etc.). Through comprehensive analysis of behavioral data and resource status, the system can reveal the key interaction relationships of customers in the business ecosystem. These relationships help identify the factors that influence customer behavior. The system generates an interaction relationship diagram between customers and other related entities based on the customer's behavioral data, such as insurance product purchase records, policy management behaviors, and interaction records with agents. For example, the interaction relationship diagram of auto insurance customers can show the interaction between customers and insurance agents, vehicle repair shops, and cooperative financial institutions. The system intuitively displays the customer's business network in the form of diagrams and provides background information for subsequent behavioral analysis.

[0128] The system identifies key factors that influence customer behavior by analyzing the interaction between customers and external entities in the interaction graph. These key features can include customer dependencies, customer resource utilization, customer response behavior, etc. After determining these features, the system can further analyze customer behavior trends in specific businesses. The system identifies the interaction frequency, interaction pattern, and resource flow between customers and external entities by analyzing the nodes and edges in the interaction graph. For example, in the insurance field, the system can analyze the interaction between customers and claims agencies, identify customers' claims behavior patterns, and the degree of cooperation dependence with external agents. These analysis results will be used in subsequent behavior prediction models.

[0129] The system inputs the key features obtained from the analysis into the previously constructed customer analysis model to predict the behavioral trends of target customers. Behavioral trend data refers to the pattern of changes in customer behavior over a specific period of time. This data can be used to assess customers' future decisions, needs, or risks. Based on the identified key customer features, such as the intensity of interaction between customers and external related entities, the frequency of resource use, etc., the system inputs these features into the customer analysis model. The customer analysis model processes these features to generate behavioral trend prediction data for target customers. For example, an auto insurance company can predict a customer's possible future insurance needs or risk levels based on data such as the customer's vehicle usage record and maintenance frequency. These prediction results can help companies provide customers with more accurate services and products.

[0130] This embodiment monitors the target customer's behavior data and resource status in real time, and the system can identify the key interactive relationships between customers and external entities, thereby extracting key features that affect customer behavior. After inputting these features into the customer analysis model, the system can predict customer behavior trends, helping companies to more accurately provide customized services or products in future business decisions, improve customer satisfaction, and reduce business risks.

[0131] In one embodiment, the above S60 includes:

[0132] S601, defining multiple business scenarios according to target business requirements, wherein the business scenarios include regular use scenarios, emergency event handling scenarios, or resource fluctuation scenarios, and generating a unique identifier for each business scenario;

[0133] S602, analyzing the time series data in each time period based on different business scenarios, and generating behavior analysis results of target customers in each time period in each business scenario;

[0134] S603, generating a corresponding behavior data set according to the behavior analysis result of each business scenario in each time period, wherein the behavior data set includes the behavior pattern, resource utilization and change trend of the target customer in each time period;

[0135] S604, integrating the behavior data set of each business scenario in each time period to form a comprehensive analysis result.

[0136] In this embodiment, a business scenario refers to a behavior pattern in a specific business situation or situation. The system defines a variety of possible scenarios based on different business needs. Common scenarios include routine usage scenarios (customer daily behavior), emergency handling scenarios (dealing with unforeseen situations), and resource fluctuation scenarios (situations with resource shortages or excess resources). These scenarios help analyze customer behavior in different environments. The system defines multiple specific scenarios based on the needs of the insurance business. For example, in the health insurance business, a routine scenario may be a customer's normal use of medical services, while an emergency scenario may involve sudden illness or emergency medical needs. Each scenario is assigned a unique identifier to ensure that the system can clearly identify the behavior pattern corresponding to each scenario during subsequent processing and analysis.

[0137] Time series data is a record of customer behavior over a specific time period. The system analyzes this data based on different business scenarios to identify customer behavior patterns in specific scenarios. The behavior in each business scenario may be different, so the system needs to perform independent analysis in each time period based on the time series data. The system segments the time series data by time period and analyzes the data for each scenario. For example, in the auto insurance scenario, the system analyzes the customer's regular driving behavior (regular scenario) and the handling behavior in unexpected accidents (emergency scenario). The system models the behavior data for each scenario and generates customer behavior analysis results for that time period.

[0138] The behavior analysis results for each business scenario in each time period include the customer's behavior patterns, resource utilization, and change trends at that specific time and scenario. These analysis results are organized into behavior data sets for more comprehensive business decisions and predictions. Based on the analyzed behavior patterns, the system generates data sets containing customer behavior data, resource utilization data, and change trends. For example, in a resource fluctuation scenario, the system analyzes the customer's response measures in the case of resource shortages and the trend of changes in their resource usage. The behavior analysis results for each time period are generated separately and archived according to different scenarios for further analysis.

[0139] The system will integrate the behavioral data sets in each business scenario in each time period to form a comprehensive analysis result. Comprehensive analysis can provide customers' overall behavioral performance in different scenarios, as well as changes in behavioral patterns over time and context. This analysis helps companies fully understand customers' behavioral trends and risk situations. The system aggregates data sets from multiple business scenarios in each time period, combines customers' overall behavioral patterns and resource utilization trends, and generates a comprehensive analysis result. For example, a health insurance company can generate a comprehensive health risk assessment report covering the entire time period based on customers' health management behaviors, emergency disease response methods, etc. This report combines health conditions in both regular and emergency scenarios to provide insurance companies with a more comprehensive customer portrait.

[0140] In this embodiment, by performing scenario analysis on time series data of different business scenarios, the system can refine the customer's behavior patterns and their changing trends in various scenarios. By integrating the behavior data sets of various business scenarios, the enterprise can obtain a more comprehensive customer behavior portrait.

[0141] In one embodiment, the above S80 includes:

[0142] S801, comparing the behavior analysis results of the target customers in each business scenario in each time period with the historical behavior data of the target customers, and analyzing the changing trend of the customer's behavior pattern and the change of resource utilization;

[0143] S802, generating a business strategy corresponding to each business scenario based on the customer's behavior pattern change trend and resource utilization change, combined with the needs of each business scenario, the business strategy including resource optimization suggestions, behavior pattern adjustment suggestions and potential risk warnings;

[0144] S803, sending the business policy of each business scenario to the business processing end.

[0145] In this embodiment, the system first obtains the behavior analysis results of the target customers in each time period and each business scenario, and compares these results with the customer's historical behavior data. Through comparison, the system can identify changes in customer behavior patterns and analyze whether these changes are abnormal or pose business risks. The system obtains customer behavior data for each time period and each business scenario from the analysis module (for example, in health insurance, the customer's health management behavior or the handling of sudden medical events). Then, the system compares this data with the customer's historical behavior data (for example, past health examination records or long-term medical expenditure trends) to identify changing trends in customer behavior patterns and changes in resource usage. These analysis results will serve as the basis for generating business strategies.

[0146] By analyzing the changing trends of customer behavior patterns, the system can determine whether the customer has new behavior patterns or deviations, and by analyzing changes in the customer's resource utilization, identify whether the customer is effectively using their resources (such as money, time, medical resources, etc.). The system identifies changes in customer behavior patterns by comparing current behavior data with historical behavior data. For example, in auto insurance, the system may find that the customer's driving behavior becomes more frequent or aggressive during a specific time period. In terms of resource utilization, the system can also monitor changes in the customer's resources (such as insurance limits) and analyze whether the customer has effectively used his resources in risk events.

[0147] The system generates corresponding business strategies based on the analyzed behavior pattern change trends and resource utilization changes, combined with the specific needs of each business scenario. Business strategies can include resource optimization suggestions, behavior adjustment suggestions, and potential risk warnings for customers. The system generates customized business strategies for different business scenarios. For example, in the field of health insurance, if a customer's health behavior pattern indicates that his health condition has deteriorated, the system can generate optimization suggestions for health management, such as increasing the frequency of physical examinations or changing health management plans. For customers who under-utilize or over-use resources, the system can generate resource optimization strategies to help customers use insurance resources more effectively.

[0148] The business strategies generated by the system usually include several parts: resource optimization suggestions are to help customers use resources more effectively in future business, behavioral pattern adjustment suggestions are to provide improvement suggestions for customers' bad behaviors in specific scenarios, and potential risk warnings are to provide early warnings of risks that customers may face in the future. The business strategies generated by the system can be optimized according to the actual situation of the customer. For example, for car insurance customers who frequently have accidents, the system will generate driving behavior adjustment suggestions and warn them of the risk of future accidents. In the field of health insurance, the system may generate resource optimization suggestions, such as suggesting that customers use medical resources more reasonably to avoid the risk of overspending on medical expenses.

[0149] The system will eventually send the business policies generated for each business scenario to the business processing end for further processing by the business department or related systems. For example, the system can pass the business policies to the relevant insurance manager or customer service platform for execution or communication with customers. The system sends the policies generated for each business scenario to the relevant business processing system through a standardized interface or API. For example, in the insurance system, the policies can be passed to the account manager's CRM system, and the account manager can communicate with the customer based on these policies to adjust the customer's premium or service content.

[0150] In this embodiment, by analyzing the changing trends of customer behavior patterns and resource utilization, the system can generate personalized strategies for specific customer business needs. These strategies include resource optimization, behavior adjustment, and potential risk warning, which can help enterprises manage customer relationships more effectively, reduce risks, and provide personalized service recommendations for customers. After sending the business strategy to the business processing end, the enterprise can quickly execute and provide feedback, thereby improving service quality and customer satisfaction.

[0151] In one embodiment, a decision-making device based on big data is provided, and the decision-making device based on big data corresponds one-to-one to the decision-making method based on big data in the above embodiment. Figure 3 , Figure 3 This is a functional module diagram of a preferred embodiment of a decision-making device based on big data of the present invention. Data collection module 10, feature extraction module 20, customer analysis model construction module 30, customer behavior analysis module 40, time series processing module 50, scenario analysis module 60, quantitative analysis module 70 and business strategy generation module 80. The functional modules are described in detail as follows:

[0152] A data collection module 10 is used to collect multi-dimensional data related to the target business, wherein the multi-dimensional data includes customer personal information, historical behavior data, resource status and / or credit record;

[0153] A feature extraction module 20 is used to extract customer features that affect business decisions from the multidimensional data, wherein the customer features include customer behavior patterns, potential problem indicators, and / or resource stability;

[0154] A customer analysis model building module 30, used to build a customer analysis model based on the customer characteristics;

[0155] A customer behavior analysis module 40 is used to obtain key features of target customers and analyze the behavior trend data of the target customers based on the key features using the customer analysis model;

[0156] A time series processing module 50 is used to process the behavior trend data of target customers in segments according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0157] A scenario analysis module 60 is used to perform scenario analysis on the time series data set in combination with multiple business scenarios to generate a behavior data set corresponding to each business scenario;

[0158] The quantitative analysis module 70 is used to perform quantitative analysis on the behavior data set corresponding to each business scenario, and generate the behavior analysis results of the target customers in each business scenario in each time period;

[0159] The business strategy generation module 80 is used to generate a target business strategy for the target customer based on the behavior analysis results of the target customer in each business scenario in each time period and in combination with the historical behavior data of the target customer.

[0160] In one embodiment, the data collection module 10 is specifically used for:

[0161] Identify multi-dimensional data sources related to the target business, the multi-dimensional data sources including customer databases, credit inquiry platforms and / or public resource data platforms;

[0162] Configuring corresponding data collection interfaces for different data sources, wherein the data collection interfaces are used to obtain customer personal information, historical behavior data, resource status and / or credit records from the data sources;

[0163] The collected data is uniformly formatted, and the data from different data sources are converted into a standardized data structure. The uniform formatting includes the unification of the time format and the conversion of the numerical unit.

[0164] In one embodiment, the feature extraction module 20 is specifically used for:

[0165] Analyze the customer's historical behavior data to determine the customer's behavior pattern, which includes the customer's usage habits, interaction frequency, and access behavior;

[0166] Analyze the customer's credit history and determine the customer's potential problem indicators, including the possibility of insufficient performance ability and abnormal behavior of the customer;

[0167] Analyze the client's resource status and determine the client's resource stability, including the client's resource sources, expenditures, and available resources;

[0168] Integrate customer behavior patterns, potential problem indicators, and resource stability to generate comprehensive customer profiles for business decision making.

[0169] In one embodiment, the customer analysis model building module 30 is specifically used to:

[0170] Determine a basic model according to the complexity of the customer characteristics and the target business requirements, wherein the basic model includes a decision tree, a support vector machine or a neural network;

[0171] Dividing the customer feature data into a training set and a validation set;

[0172] Inputting the training set of customer feature data into the basic model, performing preliminary model training, and optimizing the parameters of the basic model through iteration;

[0173] The trained model is analyzed using the validation set to determine the performance indicator results of the model on the validation set, and the hyperparameters of the model are adjusted according to the performance indicator results to generate a trained customer analysis model.

[0174] In one embodiment, the customer behavior analysis module 40 is specifically used to:

[0175] Collect target customers’ behavior data and resource status through real-time data monitoring module;

[0176] Based on the behavior data and resource status of the target customer, generating an interaction relationship diagram between the target customer and external related entities;

[0177] Based on the interaction relationship diagram, analyzing the interaction behavior between the target customer and the external related entity, and determining the key features that affect the target customer's behavior;

[0178] The key features are input into a customer analysis model to analyze the behavior trend data of the target customers, wherein the behavior trend data reflects the behavior analysis results of the target customers in multiple time periods.

[0179] In one embodiment, the scene analysis module 60 is specifically configured to:

[0180] According to the target business requirements, multiple business scenarios are defined, including regular use scenarios, emergency event handling scenarios, or resource fluctuation scenarios, and a unique identifier is generated for each business scenario;

[0181] Based on different business scenarios, analyze the time series data in each time period to generate the behavior analysis results of the target customers in each time period in each business scenario;

[0182] Generate a corresponding behavior data set based on the behavior analysis results of each business scenario in each time period, where the behavior data set includes the behavior pattern, resource utilization and change trend of the target customer in each time period;

[0183] Integrate the behavioral data sets of each business scenario in each time period to form a comprehensive analysis result.

[0184] In one embodiment, the business policy generation module 80 is specifically used to:

[0185] Compare the behavior analysis results of target customers in each business scenario in each time period with the historical behavior data of target customers to analyze the changing trends of customer behavior patterns and changes in resource utilization;

[0186] Generate business strategies for each business scenario based on the changing trends of customer behavior patterns and resource utilization, combined with the needs of each business scenario. The business strategies include resource optimization suggestions, behavior pattern adjustment suggestions, and potential risk warnings.

[0187] Send the business strategy for each business scenario to the business processing end.

[0188] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 4 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile and / or volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external user terminal through a network connection. When the computer program is executed by the processor, it realizes the functions or steps of a decision-making method based on big data on the service side.

[0189] In one embodiment, a computer device is provided. The computer device may be a user terminal, and its internal structure diagram may be as follows: Figure 5 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external server through a network connection. When the computer program is executed by the processor, it realizes the functions or steps on the user side of a decision-making method based on big data

[0190] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0191] Collect multi-dimensional data related to the target business, including customer personal information, historical behavior data, resource status and / or credit history;

[0192] Extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability;

[0193] Building a customer analysis model based on the customer characteristics;

[0194] Acquire key features of target customers, and use the customer analysis model to analyze behavioral trend data of the target customers based on the key features;

[0195] Segmenting the target customer's behavior trend data according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0196] Combine multiple business scenarios to perform scenario analysis on the time series data set to generate a corresponding behavior data set for each business scenario;

[0197] Quantitatively analyze the behavior data set corresponding to each business scenario to generate the behavior analysis results of target customers in each business scenario in each time period;

[0198] Based on the behavior analysis results of target customers in each business scenario in each time period, combined with the historical behavior data of target customers, target business strategies are generated for target customers.

[0199] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0200] Collect multi-dimensional data related to the target business, including customer personal information, historical behavior data, resource status and / or credit history;

[0201] Extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability;

[0202] Building a customer analysis model based on the customer characteristics;

[0203] Acquire key features of target customers, and use the customer analysis model to analyze behavioral trend data of the target customers based on the key features;

[0204] Segmenting the target customer's behavior trend data according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods;

[0205] Combine multiple business scenarios to perform scenario analysis on the time series data set to generate a corresponding behavior data set for each business scenario;

[0206] Quantitatively analyze the behavior data set corresponding to each business scenario to generate the behavior analysis results of target customers in each business scenario in each time period;

[0207] Based on the behavior analysis results of target customers in each business scenario in each time period, combined with the historical behavior data of target customers, target business strategies are generated for target customers.

[0208] It should be noted that the above functions or steps that can be implemented by the computer-readable storage medium or computer device can refer to the relevant descriptions on the server side and the user side in the aforementioned method embodiment. To avoid repetition, they will not be described one by one here.

[0209] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0210] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0211] It should be noted that if software tools or components other than those of the Company appear in the embodiments of the present application, they are only used for illustration and do not represent actual use. The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the above-mentioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the above-mentioned embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents; and these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A decision making method based on big data, characterized in that: The following steps are involved: Collect multi-dimensional data related to the target business, including customer personal information, historical behavior data, resource status and / or credit history; Extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability; Building a customer analysis model based on the customer characteristics; Acquire key features of target customers, and use the customer analysis model to analyze behavioral trend data of the target customers based on the key features; Segmenting the target customer's behavior trend data according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods; Combine multiple business scenarios to perform scenario analysis on the time series data set to generate a corresponding behavior data set for each business scenario; Quantitatively analyze the behavior data set corresponding to each business scenario to generate the behavior analysis results of target customers in each business scenario in each time period; Based on the behavior analysis results of target customers in each business scenario in each time period, combined with the historical behavior data of target customers, target business strategies are generated for target customers.

2. The decision-making method based on big data as claimed in claim 1, characterized in that: Collect multi-dimensional data related to the target business, including: Identify multi-dimensional data sources related to the target business, the multi-dimensional data sources including customer databases, credit inquiry platforms and / or public resource data platforms; Configuring corresponding data collection interfaces for different data sources, wherein the data collection interfaces are used to obtain customer personal information, historical behavior data, resource status and / or credit records from the data sources; The collected data is uniformly formatted, and the data from different data sources are converted into a standardized data structure. The uniform formatting includes the unification of the time format and the conversion of the numerical unit.

3. The decision-making method based on big data as claimed in claim 1, characterized in that: Extracting customer characteristics that affect business decisions from the multidimensional data, the customer characteristics including customer behavior patterns, potential problem indicators and / or resource stability, including: Analyze the customer's historical behavior data to determine the customer's behavior pattern, which includes the customer's usage habits, interaction frequency, and access behavior; Analyze the customer's credit history and determine the customer's potential problem indicators, including the possibility of insufficient performance ability and abnormal behavior of the customer; Analyze the client's resource status and determine the client's resource stability, including the client's resource sources, expenditures, and available resources; Integrate customer behavior patterns, potential problem indicators, and resource stability to generate comprehensive customer profiles for business decision making.

4. The decision making method based on big data as claimed in claim 1, characterized in that: Based on the customer characteristics, a customer analysis model is constructed, including: Determine a basic model according to the complexity of the customer characteristics and the target business requirements, wherein the basic model includes a decision tree, a support vector machine or a neural network; Dividing the customer feature data into a training set and a validation set; Inputting the training set of customer feature data into the basic model, performing preliminary model training, and optimizing the parameters of the basic model through iteration; The trained model is analyzed using the validation set to determine the performance indicator results of the model on the validation set, and the hyperparameters of the model are adjusted according to the performance indicator results to generate a trained customer analysis model.

5. The decision making method based on big data as claimed in claim 1, characterized in that: Acquire key features of target customers, and use the customer analysis model to analyze the target customers' behavior trend data based on the key features, including: Collect target customers’ behavior data and resource status through real-time data monitoring module; Based on the behavior data and resource status of the target customer, generating an interaction relationship diagram between the target customer and external related entities; Based on the interaction relationship diagram, analyzing the interaction behavior between the target customer and the external related entity, and determining the key features that affect the target customer's behavior; The key features are input into a customer analysis model to analyze the behavior trend data of the target customers, wherein the behavior trend data reflects the behavior analysis results of the target customers in multiple time periods.

6. The decision making method based on big data as claimed in claim 1, characterized in that: Combined with multiple business scenarios, scenario analysis is performed on the time series dataset to generate corresponding behavior datasets for each business scenario, including: According to the target business requirements, multiple business scenarios are defined, including regular use scenarios, emergency event handling scenarios, or resource fluctuation scenarios, and a unique identifier is generated for each business scenario; Based on different business scenarios, analyze the time series data in each time period to generate the behavior analysis results of the target customers in each time period in each business scenario; Generate a corresponding behavior data set based on the behavior analysis results of each business scenario in each time period, where the behavior data set includes the behavior pattern, resource utilization and change trend of the target customer in each time period; Integrate the behavioral data sets of each business scenario in each time period to form a comprehensive analysis result.

7. The decision making method based on big data as claimed in claim 1, characterized in that: Based on the behavior analysis results of target customers in each business scenario in each time period, combined with the historical behavior data of target customers, target business strategies are generated for target customers, including: Compare the behavior analysis results of target customers in each business scenario in each time period with the historical behavior data of target customers to analyze the changing trends of customer behavior patterns and changes in resource utilization; Generate business strategies for each business scenario based on the changing trends of customer behavior patterns and resource utilization, combined with the needs of each business scenario. The business strategies include resource optimization suggestions, behavior pattern adjustment suggestions, and potential risk warnings. Send the business strategy for each business scenario to the business processing end.

8. A decision-making device based on big data, characterized in that: The decision making device based on big data includes: A data collection module, used to collect multi-dimensional data related to the target business, the multi-dimensional data including customer personal information, historical behavior data, resource status and / or credit record; A feature extraction module, used to extract customer features that affect business decisions from the multidimensional data, wherein the customer features include customer behavior patterns, potential problem indicators and / or resource stability; A customer analysis model building module, used to build a customer analysis model based on the customer characteristics; A customer behavior analysis module, used to obtain key features of target customers and analyze the behavior trend data of the target customers based on the key features using the customer analysis model; A time series processing module is used to process the behavior trend data of target customers in segments according to time periods to generate a time series data set, wherein the time series data set includes behavior trend data in multiple time periods; A scenario analysis module, used to combine multiple business scenarios, perform scenario analysis on the time series data set, and generate a corresponding behavior data set under each business scenario; The quantitative analysis module is used to perform quantitative analysis on the corresponding behavioral data set in each business scenario and generate the behavioral analysis results of the target customers in each business scenario in each time period; The business strategy generation module is used to generate target business strategies for target customers based on the behavior analysis results of target customers in each business scenario in each time period and in combination with the historical behavior data of target customers.

9. A computer device, characterized in that: The computer device includes a memory, a processor, and a big data-based decision generation program stored in the memory and executable on the processor. When the big data-based decision generation program is executed by the processor, the steps of the big data-based decision generation method as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The storage medium stores a decision generation program based on big data, and when the decision generation program based on big data is executed by the processor, the steps of the decision generation method based on big data as described in any one of claims 1-7 are implemented.

Images