Service prediction recommendation method and device, storage medium and electronic equipment

Abstract

The invention discloses a service prediction recommendation method and device, a storage medium and electronic equipment, and the method comprises the steps: receiving a parameter file inputted by a target object, processing the parameter file through a target value analysis model, and obtaining a plurality of pieces of interaction information, the multiple pieces of interaction information comprise at least one of the following information: a first type of parameter portrait, a second type of parameter portrait, a behavior parameter corresponding to the second type of portrait, and a resource acquisition behavior corresponding to a target object of the imported parameter file; the parameter file is a service adjustment parameter configured based on the service demand of the target object; determining a service level of the target object according to the multiple pieces of interactive output information, and estimating service demand change data of the target object; and performing service prediction recommendation on the target object based on the service level, the service demand change data and the multiple pieces of interaction information. The problems that in the related technology, the evaluation efficiency of the target object is low, and the most suitable service recommendation efficiency is not high are solved.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and more specifically, to a method, apparatus, storage medium, and electronic device for predicting and recommending services. Background Technology

[0002] As financial markets become increasingly diverse and complex, the issuance and sale of financial assets face unprecedented challenges and opportunities. In this process, a precise understanding and prediction of customer behavior becomes crucial, directly impacting asset pricing, sales strategy development, and subsequent risk management. Traditional customer analysis methods often rely on manual statistics and experience-based judgment. This approach is not only time-consuming and labor-intensive but also lacks depth and breadth, making it difficult to capture subtle changes and potential trends in customer behavior, especially when dealing with large-scale datasets.

[0003] Traditional financial analysis typically uses charts to display static information, often limited to finite dimensions such as financial statements and historical transaction records. Furthermore, the statistical methods are heavily influenced by the subjective experience and intuition of business experts. For example, in the preparation stage of financial asset issuance, understanding the behavioral preferences of potential investors and assessing their purchasing intentions and capabilities is crucial, but traditional manual analysis methods struggle to provide accurate and comprehensive insights. While the development of information systems has brought convenience to data processing, existing data systems still have shortcomings in customer analysis. On the one hand, the reports and analyses provided by these systems are often limited to existing data fields and fixed analytical frameworks, lacking flexibility and customization. On the other hand, the systems have limited ability to interpret and predict data, failing to uncover deep behavioral patterns and predictive models from historical data, leaving customer analysis at a relatively superficial level.

[0004] Therefore, in related technologies, there is still no effective solution to the problems of inefficient evaluation of target objects and low efficiency in recommending the most suitable service. Summary of the Invention

[0005] This application provides a method, system, storage medium, and electronic device for predicting and recommending services, in order to at least solve the problems of low efficiency in evaluating target objects and low efficiency in recommending the most suitable service in related technologies.

[0006] According to one embodiment of this application, a method for predicting and recommending services is provided, comprising: receiving a parameter file input by a target object, and processing the parameter file through a target value analysis model to obtain multiple interaction information, wherein the multiple interaction information includes at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object imported into the parameter file; the parameter file is a service adjustment parameter configured based on the service needs of the target object; determining the service level of the target object based on the multiple interaction output information, and estimating the service demand change data of the target object; and predicting and recommending services for the target object based on the service level, the service demand change data, and the multiple interaction information.

[0007] In an exemplary embodiment, after determining the service level of the target object based on the plurality of interactive output information and estimating the service demand change data of the target object, the method further includes: acquiring historical service data of the target object and determining the estimated service level corresponding to the service demand change data; parsing the historical service data to obtain the historical service level of the target object; determining a first level change amount between the historical service level and the current service level, and determining a second level change amount between the estimated service level and the current service level; comparing the first level change amount and the second level change amount to determine whether the service demand change data is valid based on the comparison result.

[0008] In one exemplary embodiment, comparing the first level change amount and the second level change amount to determine whether the service demand change data is valid based on the comparison result includes: determining that the service demand change data is valid data when the first level change amount is greater than or equal to the second level change amount; and determining that the service demand change data is invalid data when the first level change amount is less than the second level change amount.

[0009] In an exemplary embodiment, after receiving the parameter file input by the target object, the method further includes: determining a target document template for storing the parameter file; specifying the data type of each column of data objects in the target document template; performing data preprocessing on the parameter file, mapping the preprocessed data to the target document template based on the data type to obtain a target dataset, and adding a version identifier to the target dataset.

[0010] In an exemplary embodiment, predicting and recommending services for the target object based on the service level, the service demand change data, and the multiple interaction information includes: dividing the service level according to multiple preset level ranges to obtain multiple first scores, wherein each preset level range corresponds to an evaluation score; transforming the service demand change data according to preset business logic and data constraints to obtain multiple second scores, wherein the preset business logic and data constraints are used to convert the change data into score coefficients; statistically analyzing the application frequency of the multiple interaction information to obtain multiple third scores; summarizing the first score, second score, and third score corresponding to each service to obtain multiple target total scores corresponding to multiple services; selecting the target service corresponding to the maximum total score from the multiple target total scores, and predicting and recommending the target object based on the target service.

[0011] In an exemplary embodiment, after predicting and recommending services to the target object based on the service level, the service demand change data, and the multiple interaction information, the method further includes: obtaining the number of times the target object uses the target service corresponding to the predicted recommendation; if the number of service uses is greater than a preset number, marking the target object as a key service object; if the number of service uses is less than or equal to the preset number, marking the target object as a non-key service object.

[0012] In an exemplary embodiment, after predicting and recommending services to the target object based on the service level, the service demand change data, and the plurality of interaction information, the method further includes: obtaining evaluation information of the target object's feedback on the predicted recommendation; and adjusting the plurality of interaction information according to the evaluation information.

[0013] According to one embodiment of this application, another service prediction and recommendation apparatus is provided, comprising: a receiving module, configured to receive a parameter file input by a target object, and process the parameter file through a target value analysis model to obtain multiple interaction information, wherein the multiple interaction information includes at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object that imported the parameter file; the parameter file is a service adjustment parameter configured based on the service needs of the target object; a first determining module, configured to determine the service level of the target object based on the multiple interaction output information, and to estimate the service demand change data of the target object; and a recommendation module, configured to predict and recommend services for the target object based on the service level, the service demand change data, and the multiple interaction information.

[0014] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, wherein a computer program is stored in the computer program, and the computer program is configured to perform the prediction and recommendation method of the above-mentioned service at runtime.

[0015] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the prediction and recommendation method for the aforementioned service through the computer program.

[0016] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program and a method for predicting and recommending the above-mentioned services when the computer program is executed by a processor.

[0017] In this embodiment, a parameter file input by the target object is received, and the parameter file is deeply processed through a user-customized target value analysis model. The parameter file contains service adjustment parameters configured based on the target object's service needs. The multiple interaction information generated after model processing covers both the first and second type of parameter profiles, as well as the behavioral parameters corresponding to the second type of profile and the target object's resource acquisition behavior. The customized design of the target value analysis model can dynamically adjust the analysis dimensions according to specific business scenarios, making the customer profile more realistic and possessing high flexibility and scalability. Based on the interaction information generated by the target value analysis model, the service level of the target object is determined, and the changing trend of service needs is predicted. The service level is divided based on multi-dimensional customer profile information, while the prediction of service demand change data is achieved through deep learning and analysis of historical data using an algorithm model. Then, based on the determined service level, the predicted service demand change data, and the aforementioned interaction information, service prediction and recommendations are made for the target object. The recommended service not only considers the target object's current needs but also proactively predicts future potential needs. The above technical solution solves the problems of low efficiency in evaluating target objects and low efficiency in recommending the most suitable service in related technologies. It enables in-depth analysis of customer behavior and accurate prediction of future needs, achieving precise matching between service supply and customer demand. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the hardware environment for a service prediction and recommendation method according to an embodiment of this application.

[0021] Figure 2 This is a flowchart of a service prediction and recommendation method according to an embodiment of this application;

[0022] Figure 3 This is a schematic diagram of the system architecture corresponding to the machine learning-based pre-issuance customer profiling and service prediction method for financial assets according to the embodiments of this application.

[0023] Figure 4 This is a flowchart illustrating the customer profiling analysis algorithm according to an embodiment of this application;

[0024] Figure 5 This is a flowchart illustrating the service prediction and analysis algorithm according to an embodiment of this application;

[0025] Figure 6 This is a schematic diagram of the front-end architecture according to an embodiment of this application;

[0026] Figure 7 This is a structural block diagram of a service prediction and recommendation device according to an embodiment of this application. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present application, the following description, in conjunction with the accompanying drawings of the embodiments of the present application, will clearly and completely describe the technical solutions of the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0028] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0029] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0030] As an optional implementation, the method embodiments provided in this application can be executed on a server device or a similar computing device. Taking running on a server device as an example, Figure 1 This is a hardware structure block diagram of a server device for a service prediction and recommendation method according to an embodiment of this application. Figure 1 As shown, the server device may include one or more ( Figure 1 Only one is shown in the diagram. A processor 102 (which may include, but is not limited to, a central processing unit (CPU), a microcontroller unit (MCU), or a programmable gate array (FPGA), etc.) and a memory 104 for storing data are also shown. The server device may further include a transmission device 106 for communication functions and an input / output device 108. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the server equipment described above. For example, the server equipment may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0031] The memory 104 can be used to store computer programs, such as application software programs and modules, like the computer program corresponding to the virtual machine network detection method in this embodiment. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, thus implementing the above-described method. The memory 104 may include high-speed random access memory and non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to server devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0032] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by a communication provider for the server device. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module used for wireless communication with the Internet.

[0033] This embodiment provides a service prediction and recommendation method, applied to the aforementioned server device. Figure 2 This is a flowchart of a service prediction and recommendation method according to an embodiment of this application, which includes the following steps:

[0034] Step S202: Receive the parameter file input by the target object, and process the parameter file through the target value analysis model to obtain multiple interaction information, wherein the multiple interaction information includes at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object that imported the parameter file; the parameter file is a service adjustment parameter configured based on the service requirements of the target object.

[0035] In short, a parameter file is received from the target audience (e.g., financial asset issuers or potential investors). This file contains service adjustment parameters configured to meet service requirements. These parameters may include time intervals, service costs, service types, customer preferences, etc., and are customized based on the target audience's actual needs and market environment to reflect dynamic changes under different business scenarios. Next, the parameter file is processed in depth using a target value analysis model. This model is a data-driven analysis tool that generates multiple interactive information based on the input parameter file, including but not limited to: a first-type parameter profile (e.g., the issuer's credit status, financial indicators, etc.), a second-type parameter profile (e.g., investor behavioral preferences, investment history, etc.), behavioral parameters corresponding to the second-type profile (e.g., investor bidding frequency, winning bid amount, etc.), and the target audience's resource acquisition behavior (e.g., financing methods, channel selection, etc.). This interactive information forms the basis for customer profiling and service requirement analysis, comprehensively reflecting the characteristics and needs of the target audience.

[0036] Step S204: Determine the service level of the target object based on the multiple interactive output information, and estimate the service demand change data of the target object;

[0037] Based on the interaction information generated in step S202, the analysis system can determine the service level of the target object. The service level reflects the importance and value of the target object in the eyes of the service provider, helping to segment customer groups and provide differentiated services. By analyzing the target object's historical behavioral data and market trends, the system predicts future changes in the target object's service demand. This step utilizes machine learning algorithms, such as time series analysis and cluster analysis, to predict demand trends based on existing data, providing a basis for service providers' decision-making.

[0038] Step S206: Based on the service level, the service demand change data, and the multiple interaction information, predict and recommend services for the target object.

[0039] Optionally, after determining the service level and estimating changes in service demand, service prediction and recommendations are made to the target audience based on all this information. These predictions and recommendations may include service product suggestions, service scheduling, and service cost forecasts, aiming to optimize service supply and meet the personalized and future needs of the target audience. Service prediction and recommendations are based on service levels, demand change data, and multiple customer profiles, ensuring the comprehensiveness and accuracy of the recommendations. Simultaneously, the system can dynamically adjust the recommendation strategy based on feedback from the target audience, continuously optimizing the accuracy of service predictions.

[0040] Through the above steps, a parameter file input from the target object is received, and the parameter file is deeply processed using a user-customized target value analysis model. The parameter file contains service adjustment parameters configured based on the target object's service needs. The multiple interaction information generated after model processing covers both the first and second type of parameter profiles, as well as the behavioral parameters corresponding to the second type of profile, and the target object's resource acquisition behavior. The customized design of the target value analysis model can dynamically adjust the analysis dimensions according to specific business scenarios, making the customer profile more realistic and possessing high flexibility and scalability. Based on the interaction information generated by the target value analysis model, the service level of the target object is determined, and the changing trend of service needs is predicted. The service level classification is based on multi-dimensional customer profile information, while the prediction of service demand change data is achieved through deep learning and analysis of historical data using an algorithm model. Then, based on the determined service level, the predicted service demand change data, and the aforementioned interaction information, service prediction and recommendations are made for the target object. The recommended service not only considers the target object's current needs but also proactively predicts future potential needs. This technical solution solves the problems of low efficiency in evaluating target objects and low efficiency in recommending the most suitable service in related technologies. It enables in-depth analysis of customer behavior and accurate prediction of future needs, achieving precise matching between service supply and customer demand.

[0041] In an exemplary embodiment, after determining the service level of the target object based on the plurality of interactive output information and estimating the service demand change data of the target object, the method further includes: acquiring historical service data of the target object and determining the estimated service level corresponding to the service demand change data; parsing the historical service data to obtain the historical service level of the target object; determining a first level change amount between the historical service level and the current service level, and determining a second level change amount between the estimated service level and the current service level; comparing the first level change amount and the second level change amount to determine whether the service demand change data is valid based on the comparison result.

[0042] Understandably, historical service data is used to construct a service history framework for the target object. This historical service data encompasses the target object's service usage over a past period, including service type, usage frequency, and satisfaction ratings. Subsequently, in-depth analysis of the collected historical service data yields the target object's historical service level. The determination of the historical service level is based on indicators such as customer needs, response, and service quality reflected in the historical data, representing a quantitative assessment of the quality and level of past service experience. Building upon this, the first-level change between the historical and current service levels is calculated, as well as the second-level change between the estimated and current service levels. The first-level change reflects the natural evolution trend of the target object's service level over time, while the second-level change is a prediction of future changes in the service level based on estimated service demand changes. Comparing the first and second-level changes verifies the effectiveness of the service demand change data. If the second-level change matches the first-level change or shows a reasonable trend, then the service demand change data can be considered effective, and the prediction results have high credibility. Conversely, if significant differences exist, the process of generating service demand change data needs to be re-examined to check whether the prediction bias is caused by factors such as model assumptions, data quality, or changes in the external environment, thereby ensuring the accuracy and reliability of the service prediction recommendation strategy.

[0043] In one exemplary embodiment, comparing the first level change amount and the second level change amount to determine whether the service demand change data is valid based on the comparison result includes: determining that the service demand change data is valid data when the first level change amount is greater than or equal to the second level change amount; and determining that the service demand change data is invalid data when the first level change amount is less than the second level change amount.

[0044] Optionally, the robustness of the prediction results can be verified by comparing the historical service level changes of the target object (first-level change) with the future service level changes (second-level change) derived from the prediction model. This involves ensuring a certain degree of consistency and rationality between the continuity of historical behavior patterns and the consistency of future predictions. Specifically, when the first-level change (i.e., the actual trend of past service level changes) is greater than or equal to the second-level change (the trend of service level changes predicted based on service demand change data), the service demand change data is considered valid. This is because the upward or stable trend shown in historical data aligns with the possible future direction of change indicated by the prediction model, demonstrating consistency between the model's predictions and its historical behavior patterns, thus increasing the reliability of the prediction. Conversely, if the first-level change is less than the second-level change, meaning the historical service level change trend shows a decline or no change, while the prediction model predicts a significant increase in service level, the service demand change data will be deemed invalid. Invalidity implies that the prediction results do not match the target object's past behavior, potentially indicating improper model parameter settings, data anomalies, or other exogenous factors interfering with the accuracy of the prediction. In this situation, it is necessary not only to verify the source and quality of the data, but also to debug and optimize the prediction model to eliminate potential biases and ensure the accuracy and reliability of service predictions.

[0045] In an exemplary embodiment, after receiving the parameter file input by the target object, the method further includes: determining a target document template for storing the parameter file; specifying the data type of each column of data objects in the target document template; performing data preprocessing on the parameter file, mapping the preprocessed data to the target document template based on the data type to obtain a target dataset, and adding a version identifier to the target dataset.

[0046] Optionally, to ensure the validity of the parameter file and the consistency of data processing, the concept of a target document template and a data preprocessing process are introduced. When the system receives the parameter file from the target object input, it first confirms a target document template. This template is designed for the parameter file and specifies the data type standards that each column of data objects should follow. For example, date parameters should be of type "date", and numeric parameters should be of type "int" or "double", etc. Subsequently, data preprocessing is performed on the parameter file. This is a crucial step in data analysis, designed to improve data quality and accuracy, including but not limited to data cleaning (removing null values ​​or non-compliant data), data transformation (such as converting text dates to date formats), and data validation (ensuring data conforms to business rules), ensuring that the input data for the algorithm model is reliable and standardized. After data preprocessing, the system maps these data to the target document template based on the specified data types, forming the target dataset. The mapping process ensures that the data structure matches the template. Furthermore, a version identifier is added to each target dataset, ensuring data traceability and version control.

[0047] Optional, version identifier records, such as "V1.0", "V1.1", etc., mark the generation time of the dataset, modification records, or any important change information.

[0048] By determining the target document template, performing data preprocessing and mapping, and adding version identifiers through the above embodiments, the standardization of data processing is effectively improved, ensuring data quality and consistency, and laying a solid foundation for subsequent customer profiling analysis and service demand prediction.

[0049] In an exemplary embodiment, predicting and recommending services for the target object based on the service level, the service demand change data, and the multiple interaction information includes: dividing the service level according to multiple preset level ranges to obtain multiple first scores, wherein each preset level range corresponds to an evaluation score; transforming the service demand change data according to preset business logic and data constraints to obtain multiple second scores, wherein the preset business logic and data constraints are used to convert the change data into score coefficients; statistically analyzing the application frequency of the multiple interaction information to obtain multiple third scores; summarizing the first score, second score, and third score corresponding to each service to obtain multiple target total scores corresponding to multiple services; selecting the target service corresponding to the maximum total score from the multiple target total scores, and predicting and recommending the target object based on the target service.

[0050] It's important to note that the quantitative evaluation of service levels involves defining multiple preset level ranges, each corresponding to an evaluation score. This transforms the abstract service levels into numerical forms easily processed by algorithms. This step makes comparing and analyzing different service levels intuitive and feasible. Secondly, service demand change data undergoes transformation processing based on preset business logic and data constraints, resulting in a series of second scores. These preset business logic and data constraints are designed based on professional knowledge and practical experience in the financial field, guiding the conversion of change data into score coefficients and ensuring the rationality and accuracy of the data transformation. Through this transformation, demand change data that might have been difficult to compare directly becomes scores that can be directly used for calculation, facilitating integration and analysis with other information. Next, the application frequency of multiple interaction information is statistically analyzed to derive multiple third scores. This interaction information covers various aspects such as customer profiles, behavioral preferences, and resource acquisition behaviors. By quantifying their usage frequency, the activity and importance of the target object can be reflected, thus assigning corresponding weights to each piece of information. Finally, the first, second, and third scores corresponding to each service are summarized to obtain the total target score for multiple services. This is achieved through a comprehensive scoring mechanism. The total score for each service is a holistic evaluation across three dimensions: service level, demand variability, and frequency of interaction, reflecting the service's all-around value and suitability. Finally, the service with the highest total score from the calculated pool of target scores is selected as the optimal option for prediction and recommendation. This process is similar to finding the optimal solution, ensuring that the recommended service best represents the target audience's needs, value, and preferences, thereby improving the accuracy of service recommendations and user satisfaction.

[0051] In an exemplary embodiment, after predicting and recommending services to the target object based on the service level, the service demand change data, and the multiple interaction information, the method further includes: obtaining the number of times the target object uses the target service corresponding to the predicted recommendation; if the number of service uses is greater than a preset number, marking the target object as a key service object; if the number of service uses is less than or equal to the preset number, marking the target object as a non-key service object.

[0052] Optionally, after predicting and recommending services to target objects based on service levels, changes in service demands, and multiple interaction information, the system further tracks and evaluates the frequency with which target objects actually use the recommended services to refine customer management and service strategies. Specifically, the system records and counts the actual number of times a target object uses the predicted and recommended target service, providing a real data foundation for subsequent customer classification and decision-making. The service frequency statistics are based on data sources such as system interface call records and service request logs, ensuring data accuracy and timeliness. Target objects are classified according to a preset frequency threshold. If the number of service visits exceeds this preset threshold, the system marks the target object as a key service object, meaning that the object uses the recommended service frequently and has relatively high dependence on and satisfaction with the service; this is a customer group that financial institutions need to focus on and maintain. Conversely, if the number of service visits is less than or equal to the preset threshold, the target object will be marked as a non-key service object, indicating that the object uses the recommended service less frequently, possibly due to unmet needs, poor service matching, or changes in customer preferences.

[0053] This classification mechanism helps financial institutions dynamically adjust the allocation of service resources, providing higher-level services and support to key users who use services frequently, while reassessing the needs of less frequent users to explore more suitable contact and retention strategies. This optimizes resource allocation and improves customer satisfaction. Furthermore, this feedback mechanism allows the system to continuously learn and optimize its service prediction model, improving the accuracy and effectiveness of future recommendations.

[0054] In an exemplary embodiment, after predicting and recommending services to the target object based on the service level, the service demand change data, and the plurality of interaction information, the method further includes: obtaining evaluation information of the target object's feedback on the predicted recommendation; and adjusting the plurality of interaction information according to the evaluation information.

[0055] Optionally, the system collects feedback from target users regarding the recommended services. This feedback may include aspects such as service suitability, satisfaction levels, and improvement suggestions, directly reflecting genuine customer feedback. Next, based on this feedback, the system makes targeted adjustments to several interactive information points used to generate predictive recommendations. This interactive information encompasses multi-dimensional customer profiles, trends in service demand, and past service interaction records, serving as crucial input to the predictive model. Guided by feedback, the system identifies which interactive information contributes more to service recommendations and which information may require further refinement or supplementation, thereby continuously iterating and optimizing the service prediction model. For example, if most target users report that the predicted recommendations for a particular service fail to adequately consider their specific needs or preferences, the system may increase the weighting of personalized customer preference information and adjust relevant model parameters to better reflect these needs in future recommendations. Alternatively, if a significant discrepancy is found between the estimated changes in service demand and actual usage, the processing logic for this data will be revised to ensure that the model predictions are more closely aligned with real-world applications. This closed-loop feedback mechanism, through a cycle of recommendation-feedback-adjustment, enables the continuous evolution and adaptation of the service prediction model, ensuring that financial institutions can promptly capture and respond to changes in market and customer needs, thereby improving the personalization of service recommendations and customer satisfaction.

[0056] To better understand the process of the prediction and recommendation method for the above services, the following description of the prediction and recommendation method flow for the above services is further illustrated with reference to optional embodiments, but is not intended to limit the technical solutions of the embodiments of this application.

[0057] As an optional implementation, this application proposes a method for pre-issuance customer profiling and service prediction based on machine learning. It utilizes a Vue+ElementUI+SpringBoot architecture to construct a front-end and back-end separated analysis system. Algorithmically, it fully integrates Python's data mining capabilities and the visualization capabilities of BI tools. It applies Markov state transition, K-means clustering analysis, and multinomial fitting algorithms to provide a multidimensional interactive data mining method. It visualizes the behavioral preferences of customers (issuers, sales targets) and predicts customer value and future value trends. Simultaneously, it establishes a multiple linear regression model based on the behavioral characteristics of sales targets to provide financial asset issuers with reference information and predictive rankings of the capabilities and willingness of sales targets. This constructs a software application method encompassing "application authentication—interactive parameter import—user-customized dynamic analysis model—machine learning analysis and prediction—visualization."

[0058] Optionally, the above method mainly includes two core components: 1. an interactive model, and 2. a visualization display. Specifically, in the interactive model, users can import parameter files for the corresponding modules and perform calculations on relevant data based on the user-customized model. User-customizable model parameters include time intervals, indicator conversion score points, indicator score coefficients, etc. The interactive model includes four functional modules: issuer customer profile, sales target customer profile, sales target behavior parameter analysis, and purchase behavior prediction. In the visualization display, the optional embodiment of this application uses a customer profile module based on an improved RFM customer evaluation model and the analytic hierarchy process to calculate the total score and status classification of the corresponding customers, dividing them into seven categories: general development customers, general retention customers, general value customers, important retention customers, important development customers, important retention customers, and high-value customers. At the same time, based on their importance in the previous, earlier, and current stages, and according to the basic principle of Markov state transition, a transition matrix is ​​calculated, and the state of the next stage is predicted, with the predicted state divided into ten categories.

[0059] Optionally, this application's embodiments optimize the RFM model as follows: First, in addition to the three conventional RFM indicators, several financial asset indicators are expanded and integrated into the RFM model. Issuer customer profile indicators include: the most recent issuance date, the number of issuances within one year, the issuance size within one year, the weighted average winning bid price, the benchmark interest rate spread, the ChinaBond Treasury yield curve, the average issuance size, the average issuance term in months, future principal repayment, future interest payment, the average remaining maturity, and the current debt size. Sales target customer profile indicators include: aggregated regional proximity, aggregated winning bid price, aggregated winning bid volume, aggregated self-held ratio, aggregated underwriting volume as a percentage of the underwriting syndicate, the number of participants in the syndicate, and the most recent underwriting interval. Integrating issuer customer profile indicators and sales target customer profile indicators into the RFM model has a solid theoretical foundation and empirical evidence. The theoretical basis is an extension of the customer lifetime value theory, relationship marketing theory, and investment behavior theory. The empirical evidence is based on three points: Berger & Nasr (1998) proved that in interbank business, adding transaction pattern indicators can improve the accuracy of customer value prediction by more than 30%; Dunbar (2000) found that in the selection of sales targets, the quality of past cooperation is a more important determinant than price; Geczy (2013) proved that the issuer's issuance pattern and the participation behavior of sales targets can predict the probability of future cooperation. Second, an interactive RFM model is created, where the score points and coefficients of each indicator can be dynamically adjusted. Values ​​are assigned based on the actual business needs of the user's financial assets and can be entered into an Excel parameter file, following the principles of "score point constraint 1<2<3" and "total coefficients equal to 1". The general calculation of RFM indicators is usually the arithmetic mean of three indicators. The improved RFM model uses interactively imported weights. Third, the observation window of the RFM model is optimized. The improved RFM evaluates customers not as a static evaluation from the past to the present, but as a dynamic process where a time interval can be selected as the observation window. New values ​​are assigned according to business needs by passing in the parameter file.

[0060] Optionally, a Markov chain is a stochastic process with Markov properties, where the current state depends only on the previous state and is independent of historical states. The Markov transition probability is the "probability" of an observed target in a given state i transitioning to state j within a given observation period. The client's state transition process conforms to this scenario. The observation period can be adjusted as needed. This method and apparatus are designed for Markov state prediction as follows: First, there are three transition states (state space): rising, falling, and stable. Based on the comparison of the RFM scores of the current stage and the previous stage, the customer state can be defined as rising, falling, or stable. There are nine state combinations: remaining stable, stable rising, stable falling, rising and stable, continuous rising, rising and regressing, falling and stable, falling and regressing, and continuous falling. The transition matrix is ​​a square matrix. The probabilities in this transition matrix depend on the statistical information of state changes within the "observation period". As needed, a natural year is used as the observation window, that is, the description and statistical information of two state transitions formed by the three nodes of the previous year, last year, and this year, to obtain a one-step Markov state transition probability matrix for subsequent prediction. The accuracy of the prediction results can be verified by calculating the RFM score.

[0061] Optionally, K-means clustering is a common unsupervised algorithm in machine learning. For the unlabeled task analyzed by this method and device, the basic algorithm ideas and usage can still be seen everywhere in the clustering models of most neural networks. The power of machine learning lies in feature extraction. For clustering scenarios with defined features, basic but appropriate clustering algorithms are available, and the number of clusters is determined based on maximizing the silhouette coefficient.

[0062] Optionally, the silhouette coefficient can be used as an evaluation metric for clustering. There are generally three types of evaluation metrics for clustering algorithms: external, internal, and relative. External evaluation metrics are based on a given benchmark, evaluating the clustering results against which there is already clearly categorized data for comparison. Examples include purity methods, the RAND index, and NMI, all of which require a truth label to evaluate the cluster label. Internal clustering evaluation metrics evaluate the clustering results when the true labels are unknown. A common drawback of internal clustering evaluation metrics is that they assume the categories in the dataset are clustered, such as the Dunn index, Davidson-Bolding index, and CH index. For each sales object, the classification rules and number of categories for financial assets are unknown, representing a scenario with uncertain labels. Internal evaluation metrics are suitable in this case. Compared to SSE (sum of squared errors), which only considers intra-cluster similarity, and SP (interval), which only considers inter-cluster similarity, the silhouette coefficient considers both intra-cluster and inter-cluster similarity. The silhouette coefficient ranges from -1 to 1. A value closer to 1 indicates better clustering performance, while a value closer to -1 indicates worse performance. A larger silhouette coefficient indicates tighter clustering within clusters and greater distance between clusters. An example implementation logic is as follows: cluster only sales objects that have won bids more than twice within a year. Iterate through K=2 to 7, perform K-means clustering for each K value, check the validity of the clustering results, calculate the silhouette coefficient, and finally select the K value with the largest silhouette coefficient as the optimal solution. Multiple linear regression solves the problem by considering all variables in a single expression. The willingness to purchase financial assets is the result of multiple factors, and in complex real-world scenarios, it's impossible to fit the impact of a single factor on purchasing intention. Not performing "feature selection" is also to analyze and demonstrate the true impact of each factor on the issuance and purchase behavior of financial assets while making predictions. "Feature scaling" is necessary because different features have different units of measurement, thus requiring scaling. Strictly passing through all data points is unrealistic, but it's always desirable to approximate the result by passing through as many data points as possible.

[0063] Optional, Figure 3 This is a schematic diagram of the system architecture corresponding to the machine learning-based pre-issuance customer profiling and service prediction method for financial assets according to the embodiments of this application; including: presentation layer, application layer, and data layer;

[0064] Among them, the display window facing the user is based on a front-end and back-end separation technical architecture and a lightweight Vue front-end technology component. The user realizes permission verification and algorithm through the login authentication service, data resource isolation, two-way data binding, provides good browser compatibility and operation experience, has horizontal expansion capability, uses CORS for cross-domain processing, and realizes seamless integration and flexible replacement of optional embodiments and devices and algorithm modules of this application.

[0065] The application layer implements the specific logic processing of the device and has horizontal scalability. In optional embodiments of this application, CORS is used for cross-domain processing to realize front-end and back-end interaction, and the device is integrated with Python scripts through Runtime command calls.

[0066] The data layer is the data resource guarantee for the optional embodiments and the overall device of this application, storing user information, basic elements and analysis results.

[0067] As an optional implementation method, Figure 4 This is a flowchart illustrating the customer profile analysis algorithm according to an embodiment of this application, including: receiving an Excel parameter file uploaded by a user; parsing the parameter file, calling a Python algorithm module to calculate various customer profile indicators, using the analytic hierarchy process to determine the multi-stage customer status, then calculating a Markov state transition matrix to predict the next stage customer status, then storing the predicted results in a database, and visually displaying the prediction results corresponding to different parameter files on a relevant interface.

[0068] Optional, Figure 5 This is a flowchart illustrating the service prediction analysis algorithm according to an embodiment of this application, including: receiving an Excel parameter file uploaded by a user; parsing the parameter file, calling a Python algorithm module, calculating K-means clustering based on the optimal silhouette coefficient, performing customer multiple linear regression element coefficient fitting, calculating behavioral prediction indicators based on interaction parameter information to generate a purchase intention recommendation list, storing the generated results in a database, and visually displaying the generated results corresponding to different parameter files on the relevant interface.

[0069] Optionally, the uploaded Excel parameter file may include, but is not limited to, issuer customer profile parameters, sales target customer profile parameters, issuer sales target behavior analysis parameters, and sales forecast parameters. For example: Table 1 is the issuer customer profile parameter table, Table 2 is the sales target customer profile parameter table, Table 3 is the issuer sales target behavior analysis parameter table, and Table 4 is the sales forecast parameter table. Wherein, `paramName` is the parameter name. It is a unique identifier used to indicate each parameter defined in the parameter file, used to identify and reference specific parameters in the backend processing logic. `paramValue` is the parameter value. It indicates the specific numerical value or set value, which participates in the calculation or logical judgment of the algorithm model, such as time interval, score point, or weight value. `paramType` is the parameter type. It defines the data type of the parameter value, such as integer (int), floating-point number (double), string (string), or date (date), to ensure format consistency and algorithm compatibility during data processing. `Describe` is the description. It is used to briefly explain the function or purpose of the parameter, helping users understand the role of the parameter in the model, such as the score point for average issuance size, the coefficient of the total target value, etc. `Remarks` are the remarks. Used to provide additional information or constraints when setting parameters, such as the constraint order of score points (constraint 1<2<3), the total sum of coefficients must be 1, etc., to ensure the rationality of parameter settings and the correctness of algorithm operation.

[0070] Table 1

[0071]

[0072]

[0073] Table 2

[0074]

[0075]

[0076] Table 3

[0077]

[0078] Table 4

[0079]

[0080] It should be noted that the parameter file mentioned above requires data preprocessing during the mapping to a table. This includes operations such as: setting score constraints (1 < 2 < 3), ensuring the total of 10 coefficients is 1, removing null values, ensuring the data conforms to business rules, and setting default values ​​for key parameters. The parameter mapping rule can be implemented by specifying the data type of each column when Python reads the Excel parameter file, allowing for direct mapping.

[0081] Optional, Figure 6 This is a schematic diagram of the front-end architecture according to an embodiment of this application; it includes three parts: user interface, business logic, and data structure.

[0082] It's worth noting that Vue, one of the frameworks mentioned above, is an excellent lightweight framework with fast execution speed, component-based development capabilities, two-way data binding, and reduced DOM manipulation. ElementUI is a desktop UI component framework based on Vue.js, a concise, intuitive, and powerful front-end UI component development framework that enables faster and simpler web application development. Vue is an MVVM framework that can synchronize asynchronous data from axios to the user page in real time, dividing a large HTML webpage into different components for maintenance. Each component is data-driven, providing a good interactive experience for users. SpringBoot uses the principle of "convention over configuration," greatly improving development and deployment efficiency. It uses Java annotations throughout, and SpringBoot's web components integrate the SpringMVC framework by default, making them ready to use out of the box, reducing environment setup and configuration work, and allowing users to focus on code development.

[0083] It's worth noting that a front-end / back-end separation architecture is adopted, with the front-end and back-end agreeing on an interaction interface. The front-end calls the back-end's API through this interface. The front-end only needs to focus on page styles and the parsing and rendering of dynamic data, while the back-end focuses on specific business logic, thus making development more flexible and efficient. Through routing configuration, the front-end can achieve on-demand page loading, eliminating the need to load all system resources at the beginning of the homepage loading process. The server also no longer needs to parse the front-end page, improving page interaction and user experience. Simultaneously, the front-end / back-end separation achieves high cohesion and low coupling, reducing the concurrency / load pressure on the back-end application server. The back-end can better pursue high concurrency, high availability, and high performance, while the front-end can better pursue page presentation, speed, smoothness, compatibility, and user experience.

[0084] In summary, through the above embodiments, two models were constructed based on the algorithm model. These models offer the following advantages: first, financial asset issuers and sales targets can customize parameters in the improved RFM customer profiling model, enabling customized analysis; second, based on clustering and fitting algorithms using optimal silhouette coefficients, the model supports interactive and scalable algorithms for analyzing purchasing behavior elements and predicting purchase intentions. On the application service device, the front-end and back-end separated application fully integrates Python data mining capabilities and BI platform visualization capabilities, providing multi-dimensional statistical reports and interactive data mining, empowering various types of pre-issuance services for financial assets.

[0085] Through the above description of the embodiments, those skilled in the art can clearly understand that the prediction and recommendation method for the services provided according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the related technology, can be embodied in the form of a software device. This computer software device is stored in a storage medium (such as ROM / RAM, disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the prediction and recommendation method for the services provided in the various embodiments of this application.

[0086] This embodiment also provides a service prediction and recommendation apparatus for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, hardware implementations, or a combination of software and hardware, are also possible and contemplated.

[0087] Figure 7 This is a structural block diagram of a service prediction and recommendation device according to an embodiment of this application; as shown below. Figure 7 As shown, it includes:

[0088] The receiving module 72 is used to receive a parameter file input by the target object, and process the parameter file through the target value analysis model to obtain multiple interaction information, wherein the multiple interaction information includes at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object that imports the parameter file; the parameter file is a service adjustment parameter configured based on the service requirements of the target object.

[0089] The first determining module 74 is used to determine the service level of the target object based on the multiple interactive output information, and to estimate the service demand change data of the target object;

[0090] The recommendation module 76 is used to predict and recommend services for the target object based on the service level, the service demand change data, and the multiple interaction information.

[0091] The aforementioned device receives a parameter file input from the target object and performs in-depth processing on the parameter file using a user-customized target value analysis model. The parameter file contains service adjustment parameters configured based on the target object's service needs. The multiple interaction information generated after model processing covers both the first and second type of parameter profiles, as well as the behavioral parameters corresponding to the second type of profile and the target object's resource acquisition behavior. The customized design of the target value analysis model allows for dynamic adjustment of analysis dimensions according to specific business scenarios, making customer profiles more realistic and possessing high flexibility and scalability. Based on the interaction information generated by the target value analysis model, the service level of the target object is determined, and the changing trend of service needs is predicted. The service level classification is based on multi-dimensional customer profile information, while the prediction of service demand change data is achieved through deep learning and analysis of historical data using an algorithmic model. Then, based on the determined service level, the predicted service demand change data, and the aforementioned interaction information, service prediction and recommendations are made for the target object. The recommended service not only considers the target object's current needs but also proactively predicts future potential needs. This technical solution solves the problems of low efficiency in evaluating target objects and low efficiency in recommending the most suitable service in related technologies. It enables in-depth analysis of customer behavior and accurate prediction of future needs, achieving precise matching between service supply and customer demand.

[0092] In an exemplary embodiment, the apparatus further includes: a second determining module, configured to determine the service level of the target object based on the plurality of interactive output information, and after estimating the service demand change data of the target object, acquire historical service data of the target object and determine the estimated service level corresponding to the service demand change data; parse the historical service data to obtain the historical service level of the target object; determine a first level change amount between the historical service level and the current service level, and determine a second level change amount between the estimated service level and the current service level; compare the first level change amount and the second level change amount to determine whether the service demand change data is valid based on the comparison result.

[0093] In an exemplary embodiment, the second determining module is further configured to determine the service demand change data as valid data when the change amount of the first level is greater than or equal to the change amount of the second level; and to determine the service demand change data as invalid data when the change amount of the first level is less than the change amount of the second level.

[0094] In one exemplary embodiment, the above apparatus further includes: a processing module, configured to receive a parameter file input by a target object, determine a target document template for storing the parameter file; specify the data type of each column of data objects in the target document template; perform data preprocessing on the parameter file, and map the preprocessed data to the target document template based on the data type to obtain a target dataset, and add a version identifier to the target dataset.

[0095] In an exemplary embodiment, the recommendation module is further configured to: classify the service level according to multiple preset level ranges to obtain multiple first scores, wherein each preset level range corresponds to an evaluation score; transform the service demand change data according to preset business logic and data constraints to obtain multiple second scores, wherein the preset business logic and data constraints are used to convert the change data into score coefficients; count the application frequency of the multiple interaction information to obtain multiple third scores; summarize the first score, second score, and third score corresponding to each service to obtain multiple target total scores corresponding to multiple services; select the target service corresponding to the maximum total score from the multiple target total scores, and make prediction recommendations for the target object based on the target service.

[0096] In an exemplary embodiment, the above apparatus further includes: a labeling module, configured to, after predicting and recommending services for the target object based on the service level, the service demand change data, and the plurality of interaction information, obtain the number of times the target object uses the target service corresponding to the predicted recommendation; if the number of service uses is greater than a preset number, label the target object as a key service object; if the number of service uses is less than or equal to the preset number, label the target object as a non-key service object.

[0097] In one exemplary embodiment, the above apparatus further includes: an adjustment module, configured to predict and recommend services to the target object based on the service level, the service demand change data, and the plurality of interaction information, and then obtain evaluation information of the target object's feedback on the predicted recommendation; and adjust the plurality of interaction information according to the evaluation information.

[0098] Embodiments of this application also provide a storage medium including a stored program, wherein the program, when executed, performs the prediction and recommendation method for any of the above services.

[0099] Optionally, in this embodiment, the storage medium may be configured to store program code for performing steps in the prediction and recommendation method of the above-described service.

[0100] Embodiments of this application also provide an electronic device, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the steps in any of the above method embodiments.

[0101] Embodiments of this application also provide a computer program product, including a computer program, wherein the computer program, when executed by a processor, performs the steps in any of the above method embodiments.

[0102] Optionally, the electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the processor and the input / output device is connected to the processor.

[0103] Optionally, in this embodiment, the processor can be configured to execute the steps in the prediction and recommendation method of the service via a computer program.

[0104] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical discs.

[0105] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementations, and will not be repeated here.

[0106] The collection, storage, use, processing, transmission, provision, and disclosure of financial data or user data involved in the technical solution of this application all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0107] It should be noted that in the embodiments of this application, certain software, components, models and other existing solutions in the industry may be mentioned. These should be regarded as exemplary and are only intended to illustrate the feasibility of implementing the technical solution of this application. However, it does not mean that the applicant has used or necessarily used the solution.

[0108] Obviously, those skilled in the art should understand that the modules or steps of this application described above can be implemented using general-purpose computing devices. They can be centralized on a single computing device or distributed across a network of M computing devices. Optionally, they can be implemented using computer-executable program code, thereby storing them in a storage device for execution by a computing device. In some cases, the steps shown or described can be performed in a different order than presented here, or they can be fabricated as separate integrated circuit modules, or the M modules or steps can be fabricated as a single integrated circuit module. Thus, this application is not limited to any particular combination of hardware and software.

[0109] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for predicting and recommending services, characterized in that, include: The system receives a parameter file input from a target object and processes the parameter file using a target value analysis model to obtain multiple interaction information. These multiple interaction information include at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object importing the parameter file. The parameter file consists of service adjustment parameters configured based on the service requirements of the target object. The service level of the target object is determined based on the multiple interactive output information, and the service demand change data of the target object is estimated. Based on the service level, the service demand change data, and the multiple interaction information, predictive recommendations are made for the target object.

2. The prediction and recommendation method for the service according to claim 1, characterized in that, After determining the service level of the target object based on the multiple interactive output information and estimating the service demand change data of the target object, the method further includes: Obtain historical service data of the target object and determine the estimated service level corresponding to the service demand change data; The historical service data is parsed to obtain the historical service level of the target object; Determine a first level change between the historical service level and the service level, and determine a second level change between the estimated service level and the service level; The change in the first level and the change in the second level are compared to determine whether the service demand change data is valid based on the comparison result.

3. The prediction and recommendation method for services according to claim 2, characterized in that, Comparing the change in the first level with the change in the second level, and determining the validity of the service demand change data based on the comparison result, includes: If the change in the first level is greater than or equal to the change in the second level, the service demand change data is determined to be valid data. If the change in the first level is less than the change in the second level, the service demand change data is determined to be invalid data.

4. The prediction and recommendation method for the service according to claim 1, characterized in that, After receiving the parameter file input by the target object, the method further includes: Determine the target document template for storing the parameter file; Specify the data type of each column of data objects in the target document template; Perform data preprocessing on the parameter file, and map the preprocessed data to the target document template based on the data type to obtain the target dataset, and add a version identifier to the target dataset.

5. The predictive recommendation method for the service according to claim 1, characterized in that, Based on the service level, the service demand change data, and the multiple interaction information, predictive recommendation of services for the target object includes: The service level is divided according to multiple preset level ranges to obtain multiple first scores, wherein each preset level range corresponds to an evaluation score; The service demand change data is transformed and processed according to preset business logic and data constraints to obtain multiple second scores, wherein the preset business logic and data constraints are used to convert the change data into score coefficients; The application frequency of the multiple interactive information is statistically analyzed to obtain multiple third scores; Summarize the first score, second score, and third score for each service to obtain the total score for multiple targets across multiple services; Select the target service corresponding to the highest total score from the plurality of target total scores, and make predictions and recommendations for the target object based on the target service.

6. The predictive recommendation method for the service according to claim 1, characterized in that, After predicting and recommending services for the target object based on the service level, the service demand change data, and the multiple interaction information, the method further includes: Obtain the number of times the target object uses the predicted and recommended target service; If the number of service requests exceeds a preset number, the target object will be marked as a key service object; If the number of service visits is less than or equal to a preset number, the target object will be marked as a non-priority service object.

7. The predictive recommendation method for the service according to claim 1, characterized in that, After predicting and recommending services for the target object based on the service level, the service demand change data, and the multiple interaction information, the method further includes: Obtain the evaluation information of the target object on the predicted recommendation feedback; The multiple interactive information is adjusted based on the evaluation information.

8. A method for predicting and recommending services, characterized in that, include: The receiving module is used to receive a parameter file input by the target object, and process the parameter file through a target value analysis model to obtain multiple interaction information, wherein the multiple interaction information includes at least one of the following: a first type of parameter profile, a second type of parameter profile, behavioral parameters corresponding to the second type of profile, and resource acquisition behavior corresponding to the target object that imported the parameter file; the parameter file is a service adjustment parameter configured based on the service requirements of the target object. The first determining module is used to determine the service level of the target object based on the multiple interactive output information, and to estimate the service demand change data of the target object; The recommendation module is used to predict and recommend services for the target object based on the service level, the service demand change data, and the multiple interaction information.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, wherein when the computer program is executed by a processor, it implements the steps of the predictive recommendation method for the service of any one of claims 1 to 7.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the predictive recommendation method for the service according to any one of claims 1 to 7.

Images