Resource prediction and personalized service recommendation method and device based on large model

Abstract

The invention provides a resource prediction and personalized service recommendation method and device based on a large model. The method comprises the steps of obtaining at least one type of historical business data; training a resource prediction model and a user analysis model by using the at least one type of historical business data; acquiring at least one type of real-time service data generated by the current operation behavior of at least one user; analyzing the real-time service data by using the resource prediction model, and outputting early warning information corresponding to each user; generating a personalized service scheme corresponding to each user based on the early warning information and the real-time business data by using a user analysis model; and recommending the corresponding personalized service scheme to each user. According to the invention, human intervention can be reduced, service processing is intelligent and automatic, timely response can be realized, user experience can be improved, and potential risks can be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of personalized service recommendation, and in particular to a resource prediction and personalized service recommendation method and device based on a large model. BACKGROUND

[0002] With the rapid development of information technology, emerging technologies such as cloud computing, cloud platforms, distributed clouds, big data, and artificial intelligence are widely used in various enterprises, institutions, and organizations. The popularity of the above technologies has greatly promoted the informatization level of various industries, and at the same time has put forward higher requirements for the dynamic scheduling management capability of resources and the personalized needs of users.

[0003] The existing resource management method mainly relies on manual monitoring and static rule setting, which is difficult to cope with dynamic changes in resource demand and complex system environment. Therefore, under this management method, how to provide personalized services for users has become a technical problem that needs to be solved in the current field of intelligent information services. SUMMARY

[0004] The present application provides a resource prediction and personalized service recommendation method and device based on a large model to at least partially solve the above problems.

[0005] In a first aspect, the present application provides a resource prediction and personalized service recommendation method based on a large model, comprising:

[0006] Obtaining at least one type of historical business data;

[0007] Training a resource prediction model and a user analysis model using the at least one type of historical business data;

[0008] Obtaining at least one type of real-time business data generated by the current operation behavior of at least one user;

[0009] Analyzing the real-time business data using the resource prediction model, and outputting a warning information corresponding to each user;

[0010] Generating a personalized service scheme corresponding to each user based on the warning information and the real-time business data using the user analysis model;

[0011] Recommending the personalized service scheme corresponding to each user to each user.

[0012] Preferably,

[0013] The analysis of the real-time business data using the resource prediction model and the output of the warning information corresponding to each user comprises:

[0014] For each user, performing:

[0015] predicting, by using the resource prediction model, a state of at least one target object based on real-time service data input, and obtaining a state prediction result of each target object, wherein the target object comprises at least one of resource capacity, resource usage peak, and service operation condition;

[0016] For each target object, when the state prediction result of the target object exceeds a preset threshold, generating early warning information based on the state prediction result, and outputting the early warning information.

[0017] Preferably,

[0018] The generating, by using the user analysis model, of a personalized service scheme corresponding to each user based on the early warning information and the real-time service data comprises:

[0019] For each user, the following is performed:

[0020] Determining, by using the user analysis model, an early warning strategy that matches the early warning information corresponding to the user;

[0021] Analyzing, by using the user analysis model, historical operation behavior of the user and the real-time service data, and identifying a commonly used resource type and a personalized demand of the user;

[0022] Generating, by using the user analysis model, a personalized service scheme that matches the commonly used resource type and the personalized demand of the user based on the early warning strategy.

[0023] Preferably,

[0024] After the analyzing, by using the resource prediction model, of the real-time service data and the outputting of the early warning information corresponding to each user, the following is further included:

[0025] Pushing the early warning information to an external person, wherein the pushing manner of the early warning information comprises at least one of short message pushing, email pushing, voice call pushing, and platform system message pushing.

[0026] Preferably,

[0027] The training, by using the at least one type of historical service data, of the resource prediction model and the user analysis model comprises:

[0028] Feature extraction is performed on each type of historical service data to obtain a feature data set.

[0029] From the feature data set, a resource prediction model data set and a user analysis model data set are screened out;

[0030] The resource prediction model is trained by using the resource prediction model data set;

[0031] training a user analysis model using the user analysis model dataset.

[0032] Preferably,

[0033] training a resource prediction model using the resource prediction model dataset, comprises:

[0034] dividing the resource prediction model dataset into a resource prediction model training set and a resource prediction model test set, and performing the following operations:

[0035] Step A1: training a resource prediction model using the resource prediction model training set;

[0036] Step B1: verifying whether the accuracy of the resource prediction model is higher than a preset threshold using the resource prediction model test set, if yes, performing Step C1, otherwise, performing Step D1;

[0037] Step C1: training a user analysis model using the user analysis model dataset;

[0038] Step D1: adjusting the model parameters of the resource prediction model, and performing Step A1.

[0039] Preferably,

[0040] training a user analysis model using the user analysis model dataset, comprises:

[0041] dividing the user analysis model dataset into a user analysis model training set and a user analysis model test set, and performing the following operations:

[0042] Step A2: training a user analysis model using the user analysis model training set;

[0043] Step B2: verifying whether the accuracy of the user analysis model is higher than a preset threshold using the user analysis model test set, if yes, performing Step C2, otherwise, performing Step D2;

[0044] Step C2: obtaining at least one type of real-time business data generated by the current operation behavior of at least one user;

[0045] Step D2: adjusting the model parameters of the user analysis model, and performing Step A2.

[0046] Preferably,

[0047] the obtaining at least one type of historical business data, comprises:

[0048] collecting at least one type of historical data stored;

[0049] Data processing is performed on each type of historical data to obtain corresponding historical service data.

[0050] Preferably,

[0051] The at least one type of real-time service data generated by the current operation behavior of the at least one user includes:

[0052] At least one type of real-time data generated by the current operation behavior of at least one user is collected.

[0053] For each user, data processing is performed on each type of real-time data collected to obtain corresponding real-time service data.

[0054] In a second aspect, the present application provides a method and device for resource prediction and personalized service recommendation based on a large model, comprising:

[0055] A data acquisition module is configured to acquire at least one type of historical service data and at least one type of real-time service data generated by the current operation behavior of at least one user.

[0056] A model training module is configured to train a resource prediction model and a user analysis model using the at least one type of historical service data acquired by the data acquisition module.

[0057] A model calling module is configured to analyze the real-time service data using the resource prediction model trained by the model training module, output early warning information corresponding to each user, and generate a personalized service scheme corresponding to each user based on the early warning information and the real-time service data acquired by the data acquisition module using the user analysis model.

[0058] A scheme recommendation module is configured to recommend to each user the personalized service scheme generated by the model calling module corresponding to the user.

[0059] In a third aspect, the present application provides a computer program product comprising computer programs / instructions, wherein the computer programs / instructions are executed by a processor to implement the method steps of any one of the first aspect.

[0060] In a fourth aspect, the present application provides a computer storage medium having a computer program stored thereon, wherein the program is executed by a processor to implement the method of any one of the first aspect.

[0061] The embodiment of the application provides a method and device for resource prediction and personalized service recommendation based on a large model. The resource prediction model and the user analysis model are trained through historical business data, and dynamic monitoring and analysis of user states are realized in combination with business data generated by real-time operation behaviors of users. The resource prediction model can timely output early warning information of users, and then the user analysis model generates a personalized service scheme on the basis of the early warning information and real-time data, and finally the personalized service scheme is recommended to the users, so that accurate response to user demand is realized. The application can reduce human intervention, make business processing intelligent and automatic, and also can respond in time, which helps to improve user experience and reduce potential risks. BRIEF DESCRIPTION OF DRAWINGS

[0062] Figure 1 is a flowchart of a method for resource prediction and personalized service recommendation based on a large model provided by an embodiment of the application;

[0063] Figure 2 is a flowchart of another method for resource prediction and personalized service recommendation based on a large model provided by an embodiment of the application;

[0064] Figure 3 is a schematic diagram of a device for resource prediction and personalized service recommendation based on a large model provided by an embodiment of the application. DETAILED DESCRIPTION

[0065] To make the purpose, technical scheme and advantages of the embodiments of the application clearer, the technical scheme in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are some embodiments of the application, rather than all the embodiments of the application. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the protection scope of the application.

[0066] Figure 1 The method for resource prediction and personalized service recommendation based on a large model provided by an embodiment of the application specifically includes the following steps:

[0067] Step 101: Obtain at least one type of historical business data;

[0068] Step 102: Train a resource prediction model and a user analysis model by using at least one type of historical business data;

[0069] Step 103: Obtain at least one type of real-time business data generated by current operation behaviors of at least one user;

[0070] Step 104: Analyze the real-time business data by using the resource prediction model, and output early warning information corresponding to each user;

[0071] Step 105: generating a personalized service scheme corresponding to each user based on the early warning information and real-time business data by using the user analysis model;

[0072] Step 106: recommending the personalized service scheme corresponding to each user to the user.

[0073] In the embodiments of the present application, the resource prediction model and the user analysis model are trained by historical business data, and the dynamic monitoring and analysis of the user state are realized by combining the business data generated by the real-time operation behavior of the user. The resource prediction model can output the early warning information of the user in a timely manner, and then the user analysis model generates a personalized service scheme based on the early warning information and real-time data, and finally recommends the service scheme to the user, so as to realize the accurate response to the user demand. The present application can reduce human intervention, make the business processing intelligent and automatic, and also can respond in a timely manner, which helps to improve the user experience and reduce potential risks.

[0074] In an embodiment of the present application, the resource prediction model is used to analyze the real-time business data, and output the early warning information corresponding to each user, including:

[0075] For each user, the following is performed:

[0076] The resource prediction model is used to predict the state of at least one target object based on the real-time business data input, and obtain the state prediction result of each target object, wherein the target object includes at least one of resource capacity, resource use peak value and business running condition;

[0077] For each target object, when the state prediction result of the target object exceeds a preset threshold, the early warning information is generated based on the state prediction result, and the early warning information is output.

[0078] In the embodiments of the present application, the resource prediction model is used to predict the state of the target object based on the real-time business data, which realizes the accurate monitoring of key indicators such as resource capacity, resource use peak value and business running condition. When the prediction result exceeds the preset threshold, the early warning information is generated and output in a timely manner. This method not only improves the ability to identify potential risks in advance, but also dynamically adjusts resource allocation and optimizes business processes, thereby improving the stability, reliability and resource utilization efficiency of the system, and providing more intelligent and efficient services for users.

[0079] In an embodiment of the present application, the user analysis model is used to generate a personalized service scheme corresponding to each user based on the early warning information and real-time business data, including:

[0080] For each user, the following is performed:

[0081] determine a pre-warning strategy matched with the pre-warning information corresponding to the user by using the user analysis model;

[0082] analyze the historical operation behavior and real-time business data of the user by using the user analysis model, and identify the commonly used resource type and personalized demand of the user;

[0083] generate a personalized service scheme matched with the commonly used resource type and personalized demand of the user for the user based on the pre-warning strategy by using the user analysis model.

[0084] In the embodiments of the present application, by using the user analysis model, the pre-warning information can be accurately matched with the corresponding pre-warning strategy, and the historical operation behavior and real-time business data of the user are comprehensively analyzed, so as to identify the commonly used resource type and personalized demand of the user. On this basis, a personalized service scheme highly matched with the demand of the user is generated based on the pre-warning strategy. This not only realizes the intelligentization and personalization of pre-warning response and service recommendation, but also effectively improves the resource utilization rate and service quality, and enhances the user experience.

[0085] In an embodiment of the present application, after analyzing the real-time business data by using the resource prediction model and outputting the pre-warning information corresponding to each user, further comprising:

[0086] push the pre-warning information to external personnel, wherein the pushing mode of the pre-warning information includes at least one of the following: short message pushing, email pushing, voice telephone pushing and platform system message pushing.

[0087] In the embodiments of the present application, the pre-warning information is timely delivered to external personnel through various pushing modes such as short message pushing, email pushing, voice telephone pushing and platform system message pushing. This multi-channel pushing mode ensures that the pre-warning information can be received by relevant personnel in the first time, so as to realize rapid response and decision-making, and avoid the expansion of potential risks. Through the flexible application of different pushing modes, more personalized services can be provided according to the actual needs and preferences of external personnel, further improving the communication efficiency and practicality of the pre-warning information.

[0088] In an embodiment of the present application, at least one type of historical business data is used to train the resource prediction model and the user analysis model, comprising:

[0089] feature extraction is performed on various types of historical business data to obtain a feature data set.

[0090] From the feature data set, a resource prediction model data set and a user analysis model data set are selected;

[0091] The resource prediction model is trained by using the resource prediction model data set;

[0092] The user analysis model is trained by using the user analysis model dataset.

[0093] In the embodiments of the present application, by comprehensively extracting features from various historical business data, a feature dataset is generated, from which a model dataset suitable for resource prediction and user analysis is screened out. This process can ensure that the training basis of the resource prediction model and the user analysis model is solid and the data is accurate and reliable. Through the training of the resource prediction model dataset, accurate prediction of resource usage can be achieved, and the scientificity and optimization degree of resource allocation are improved. Through the training of the user analysis model dataset, the behavior characteristics and needs of users can be deeply understood, and personalized services and optimization schemes can be provided.

[0094] In an embodiment of the present application, the resource prediction model is trained by using the resource prediction model dataset, comprising:

[0095] The resource prediction model dataset is divided into a resource prediction model training set and a resource prediction model test set, and the following operations are performed:

[0096] Step A1: training the resource prediction model by using the resource prediction model training set;

[0097] Step B1: verifying whether the accuracy of the resource prediction model is higher than a preset threshold by using the resource prediction model test set, if yes, executing step C1, otherwise, executing step D1;

[0098] Step C1: training the user analysis model by using the user analysis model dataset;

[0099] Step D1: adjusting the model parameters of the resource prediction model, and executing step A1.

[0100] In the embodiments of the present application, by dividing the resource prediction model dataset into a training set and a test set, a strict model training and verification process is implemented, ensuring the high precision and reliability of the resource prediction model. In the training phase, the model is trained by using the resource prediction model training set, which can accurately capture the rules and trends in the business data; in the test phase, whether the accuracy of the resource prediction model meets the preset threshold is verified, ensuring the usability and accuracy of the prediction result. When the model reaches the expected precision, another model can be trained; if the accuracy does not meet the requirements, the model parameters are automatically adjusted and retrained until the best prediction effect is achieved. This method can dynamically optimize the resource prediction model, effectively improving the prediction accuracy and system adaptability, thereby ensuring the accuracy of resource allocation and the timeliness of business response.

[0101] It should be noted that the resource prediction model can be trained first, and if the accuracy of the model is higher than the preset threshold, the user analysis model is trained again. The user analysis model can also be trained first, and if the accuracy of the model is higher than the preset threshold, the resource prediction model is trained again. That is, the training of the two models does not have a priority.

[0102] In an embodiment of the present application, the user analysis model is trained using the user analysis model dataset, comprising:

[0103] The user analysis model dataset is divided into a user analysis model training set and a user analysis model test set, and the following operations are performed:

[0104] Step A2: training the user analysis model using the user analysis model training set;

[0105] Step B2: verifying whether the accuracy of the user analysis model is higher than the preset threshold using the user analysis model test set, if yes, performing step C2, otherwise, performing step D2;

[0106] Step C2: obtaining at least one type of real-time business data generated by the current operation behavior of at least one user;

[0107] Step D2: adjusting the model parameters of the user analysis model, and performing step A2.

[0108] In an embodiment of the present application, by dividing the user analysis model dataset into a training set and a test set, precise model training and verification are implemented to ensure that the user analysis model can accurately reflect user behavior and demand. In the training phase, the model is trained by using the user analysis model training set, which can effectively learn the characteristics and patterns of user behavior; in the test phase, the accuracy of the user analysis model is verified to ensure that it meets the preset threshold, thereby ensuring the reliability and practical application value of the model. If the accuracy meets the requirements, accurate user analysis and personalized services can be provided based on the real-time operation behavior of the user; if the accuracy is insufficient, the model parameters are automatically adjusted for retraining until the desired effect is achieved. This method effectively improves the accuracy and real-time performance of user behavior analysis, thereby providing strong support for personalized recommendation, precise marketing and service optimization, and significantly improving user experience.

[0109] It should be noted that the resource prediction model and the user analysis model are obtained based on a large model and deep learning training.

[0110] In an embodiment of the present application, at least one type of historical business data is obtained, comprising:

[0111] At least one type of historical data stored is collected;

[0112] Each type of historical data is processed to obtain corresponding historical business data.

[0113] In the embodiment of the present application, at least one type of historical data stored is collected, and different types of historical data are processed to obtain corresponding historical business data, thereby ensuring the completeness and standardization of data input. This method not only effectively eliminates redundancy and noise in the original data, improves the accuracy and usability of the data, but also makes the obtained historical business data more representative and valuable for analysis.

[0114] It should be noted that data processing includes data cleaning, data conversion, data integration, data filtering, data aggregation, data mining and analysis, data visualization, etc., but is not limited thereto.

[0115] In an embodiment of the present application, at least one type of real-time business data generated by the current operation behavior of at least one user is obtained, including:

[0116] Collecting at least one type of real-time data generated by the current operation behavior of at least one user;

[0117] For each user, the collected real-time data of each type is processed to obtain corresponding real-time business data.

[0118] In the embodiment of the present application, at least one type of real-time data generated by the current operation behavior of at least one user is collected, and the collected real-time data is processed to obtain corresponding real-time business data, thereby realizing the instant linkage of user behavior and business processing, significantly improving the response speed and processing accuracy of user operation, ensuring the high timeliness and dynamic updating of business data, and helping to discover user demand and behavior pattern changes in time.

[0119] As shown in Figure 2 To more clearly illustrate the technical solutions and advantages of the present application, the method for resource prediction and personalized service recommendation based on a large model provided in the embodiments of the present application is described in detail below, which can specifically include the following steps:

[0120] Step 201: Collecting at least one type of historical data stored. Each type of historical data is processed to obtain corresponding historical business data.

[0121] Specifically, by collecting historical data, past business or user behavior data is utilized to provide a foundation for subsequent data analysis and model training. These data can help identify and understand past user behavior patterns, trends, and the course of business development. Data processing can clean, transform, and format historical data to meet business analysis needs. By cleaning up redundancy, noise, and missing values, the quality and accuracy of the data are ensured, and the credibility of subsequent analysis is improved. After data processing, historical data becomes more standardized, facilitating further modeling, analysis, and decision support.

[0122] For example, the collected data includes resource monitoring data, operation order processing data, user portrait data, and transaction order data.

[0123] Step 202: Feature extraction is performed on various historical business data to obtain a feature dataset. From the feature dataset, a resource prediction model dataset and a user analysis model dataset are selected.

[0124] Specifically, through feature extraction, key factors or important features can be extracted from raw business data, which can better reflect the essence of business operation. The extracted features help to train the model more effectively and help the system discover potential rules or trends from complex business data.

[0125] Specifically, the feature dataset is divided into different subsets, so that each model can focus on a specific task. The resource prediction model focuses on the change of resource demand, while the user analysis model focuses on the analysis and prediction of user behavior. Through this grouping, the model can ensure that it can handle its own task specifically, improving efficiency and prediction accuracy.

[0126] Step 203: The resource prediction model dataset is divided into a resource prediction model training set and a resource prediction model test set.

[0127] Specifically, the segmentation of the data set is a key step in model training. By dividing the data into training and test sets, it can be ensured that the model will not overfit during training and can effectively predict unseen data. In this way, the trained resource prediction model can have good generalization ability, ensuring its stability in actual application.

[0128] It can be understood that the resource prediction model training set and the resource prediction model test set can be randomly divided, or can be divided by time, without being limited thereto.

[0129] Step 204: The resource prediction model is trained using the resource prediction model training set.

[0130] Step 205: Verify whether the accuracy of the resource prediction model is higher than the preset threshold using the resource prediction model test set. If yes, execute step 207; otherwise, execute step 206.

[0131] Specifically, the resource prediction model is trained by the resource prediction model training set, so that it learns to extract rules from historical business data and establishes an effective prediction mechanism. This process enables the model to automatically identify which features have a significant impact on resource demand, thereby improving its prediction accuracy. By verifying the accuracy of the resource prediction model, it can be ensured that it has sufficient prediction ability. If the accuracy of the resource prediction model does not meet the requirements, it is adjusted or retrained to ensure that the final resource prediction model has high accuracy and reliability.

[0132] Step 206: Adjust the model parameters of the resource prediction model and execute step 204.

[0133] Specifically, when the accuracy of the resource prediction model does not meet the requirements, adjusting the model parameters can improve the performance and accuracy of the resource prediction model. The tuning process of the resource prediction model can optimize the algorithm according to the actual situation, so that the resource prediction model can make more accurate predictions when facing different types of business data.

[0134] Step 207: Divide the user analysis model dataset into a user analysis model training set and a user analysis model test set.

[0135] Specifically, by dividing the dataset into a training set and a test set, it ensures that the user analysis model can learn effective user behavior patterns from a large amount of data, and verify its prediction accuracy through the test set to prevent overfitting, ensuring that it can handle unseen data in actual applications.

[0136] Step 208: Train the user analysis model using the user analysis model training set.

[0137] Specifically, the user analysis model is trained by the user analysis model training set, so that the user analysis model can identify the potential patterns of user behavior and provide personalized predictions and recommendations. This step can help enterprises understand user preferences, demand changes, etc., thereby achieving precise marketing or product recommendations.

[0138] Step 209: Verify whether the accuracy of the user analysis model is higher than the preset threshold using the user analysis model test set. If yes, execute step 211; otherwise, execute step 210.

[0139] Specifically, the accuracy of the user analysis model is verified by the user analysis model test set to ensure that it can work effectively in actual business. If the accuracy is not high, adjust the model parameters for optimization to improve the prediction ability of the model and the effect of business application.

[0140] Step 210: Adjust the model parameters of the user analysis model, and execute step 208.

[0141] Specifically, the parameters of the user analysis model are adjusted so that the model can better function in various business scenarios and improve prediction accuracy. Optimization of the model can help the system provide services that better meet user needs and enhance user experience.

[0142] Step 211: Collect at least one type of real-time data generated by the current operation behavior of at least one user. For each user, process each type of real-time data collected to obtain corresponding real-time business data.

[0143] Specifically, the collection of real-time data ensures that the system can respond to user behavior and needs in real time, maintaining the dynamic adaptability of the business. Collecting real-time data supports subsequent real-time business processing and helps to adjust services and decisions in real time during user operations. Through the processing of real-time data, the integrity, accuracy and timeliness of the data can be ensured, providing support for subsequent business decision-making and recommendations. The generation and application of real-time business data help to improve the response speed and accuracy of user behavior.

[0144] Step 212: For each user, use the resource prediction model to predict the state of at least one target object based on real-time business data input, and obtain the state prediction result of each target object, wherein the target object includes at least one of resource capacity, resource usage peak, and business operation situation; for each target object, when the state prediction result of the target object exceeds a preset threshold, generate a warning information based on the state prediction result, and output the warning information.

[0145] Specifically, by using the resource prediction model to analyze real-time business data, various target objects (such as resource capacity, resource usage peak, and business operation situation) can be dynamically monitored, and warning information can be automatically generated when the prediction result exceeds the preset threshold. This mechanism can greatly improve the response speed and reliability of the business, ensuring that measures are taken in a timely manner before potential problems occur, thereby avoiding business interruption or resource bottlenecks. By taking preventive measures in advance, the system can optimize resource allocation, improve resource utilization efficiency, support intelligent decision-making, and help managers make data-based optimization adjustments. This prediction and warning mechanism not only enhances the stability and flexibility of the system, but also significantly improves user experience, ensures business continuity, and reduces customer loss due to system failures or performance problems.

[0146] In addition, after outputting the early warning information, the early warning information can also be pushed to external personnel. The specific pushing methods include at least one of short message, email, voice call and platform system message. This operation can ensure that important information is timely and effectively conveyed in different scenarios. This multi-channel pushing method can cover different types of users and scenarios, ensuring that the early warning information can be quickly conveyed to relevant personnel and will not be delayed due to the failure of a single channel. No matter where the external personnel are located or which information receiving method they prefer, timely delivery of information can be ensured, thereby speeding up the response speed and avoiding further deterioration of potential problems. At the same time, this flexible pushing method improves the real-time and effectiveness of early warning, which helps to take measures more quickly, reduces business risks, enhances the stability and reliability of the system, and ultimately improves the trust and satisfaction of users and relevant personnel.

[0147] Step 213: For each user, determine the early warning strategy that matches the early warning information corresponding to the user using the user analysis model; analyze the user's historical operation behavior and real-time business data using the user analysis model to identify the user's commonly used resource type and personalized needs; and generate a personalized service plan that matches the commonly used resource type and personalized needs for the user based on the early warning strategy using the user analysis model.

[0148] Step 214: Recommend the personalized service plan corresponding to each user to the user.

[0149] Specifically, by combining the user analysis model, the commonly used resource type and personalized needs of each user can be accurately identified based on their historical operation behavior and real-time business data, and matching early warning strategies and personalized service plans can be developed for users based on this information. This method not only improves user experience and makes services more tailored to individual needs, but also enables more efficient personalized customization in resource management and service provision. By customizing early warning strategies and service plans for users, user needs can be responded to more promptly, resource allocation can be optimized, resource waste can be avoided, and overall flexibility and response capabilities can be improved. In addition, this personalized service plan can improve customer satisfaction and enhance user stickiness, thereby improving the long-term stability and profitability of business operations.

[0150] It should be noted that the front end of the present application is implemented based on the Angular framework (such as obtaining real-time business data of users, interacting with users, recommending personalized plans, etc.). The back end is implemented based on the SpringBoot framework (such as storing historical business data, training resource prediction models and user analysis models, data analysis, generating personalized service plans, etc.)

[0151] As shown in Figure 3 The present application provides a device for resource prediction and personalized service recommendation based on a large model, which includes:

[0152] The data acquisition module 301 is configured to acquire at least one type of historical service data and at least one type of real-time service data generated by current operation behavior of at least one user.

[0153] The model training module 302 is configured to train a resource prediction model and a user analysis model by using the at least one type of historical service data acquired by the data acquisition module 301.

[0154] The model calling module 303 is configured to analyze the real-time service data by using the resource prediction model trained by the model training module 302, and output early warning information corresponding to each user; and generate a personalized service scheme corresponding to each user based on the early warning information and the real-time service data acquired by the data acquisition module 301 by using the user analysis model.

[0155] The scheme recommendation module 304 is configured to recommend the personalized service scheme generated by the model calling module 303 to each user.

[0156] In an embodiment of the present application, the resource prediction model and the user analysis model are trained by using historical service data, and dynamic monitoring and analysis of user states are realized by combining service data generated by real-time operation behavior of the user. The resource prediction model can output early warning information of the user in a timely manner, and then the user analysis model generates a personalized service scheme based on the early warning information and real-time data, and finally recommends the personalized service scheme to the user, so as to realize accurate response to user demand. The present application can reduce human intervention, make business processing intelligent and automatic, and also can respond in a timely manner, which is helpful to improve user experience and reduce potential risks.

[0157] Embodiments of the present application provide an electronic device, and embodiments of the present application do not limit the specific implementation of the electronic device.

[0158] The electronic device can include a processor, a communications interface, a memory, and a communications bus. Wherein:

[0159] The processor, the communications interface, and the memory complete mutual communication through the communications bus.

[0160] The communications interface is configured to communicate with other electronic devices or servers.

[0161] The processor is configured to execute a program, and specifically can execute related steps in the above method embodiments.

[0162] Specifically, the program can include program code, and the program code includes computer operation instructions.

[0163] The processor can be a CPU, or a GPU (Graphic Processing Unit) or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement one or more embodiments of the present application. The one or more processors included in the smart device can be processors of the same type, such as one or more CPUs; or can be processors of different types, such as one or more CPUs and one or more ASICs.

[0164] The memory is configured to store programs. The memory can include a high-speed RAM memory, and can further include a non-volatile memory, such as at least one disk memory.

[0165] The program can include a plurality of computer instructions, and the program can specifically cause the processor to perform operations corresponding to the method described in any one of the foregoing method embodiments.

[0166] The specific implementation of each step in the program can refer to the corresponding description in the corresponding steps and units in the foregoing method embodiments, and has corresponding beneficial effects, which will not be described here. Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the device and the module described above can refer to the corresponding process description in the foregoing method embodiments, which will not be described here

[0167] The embodiments of the present application further provide a computer storage medium, which stores a computer program, and the program is executed by a processor to implement the method described in any one of the foregoing method embodiments. The computer storage medium includes but is not limited to: a CD-ROM (Compact Disc Read-Only Memory), a RAM (Random Access Memory), a floppy disk, a hard disk or a magneto-optical disk, etc.

[0168] The embodiments of the present application further provide a computer program product, which includes computer instructions, and the computer instructions instruct a computing device to perform operations corresponding to any one of the foregoing method embodiments.

[0169] In addition, it should be noted that the information related to the user (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to sample data for training the model, data for analysis, stored data, displayed data, etc.) involved in the embodiments of the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of the relevant data need to comply with the relevant laws, regulations and standards of the country and region, and provide corresponding operation portal for the user to choose authorization or refusal.

[0170] It should be noted that, according to the needs of implementation, each component / step described in the embodiments of the present application can be split into more components / steps, or two or more components / steps or part of the operation of the components / steps can be combined into a new component / step, to achieve the purpose of the embodiments of the present application.

[0171] The above method according to the embodiments of the present application can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium such as a CD-ROM, RAM, floppy disk, hard disk or magneto-optical disk, or be implemented by computer code originally stored in a remote recording medium or non-transitory machine readable medium downloaded through a network and stored in a local recording medium, so that the method described herein can be stored in such software processing of recording medium using a general computer, special processor or programmable or special hardware such as Application Specific Integrated Circuit (ASIC) or Field Programmable Gate Array (FPGA). It can be understood that the computer, processor, microprocessor controller or programmable hardware includes storage components (for example, Random Access Memory (RAM), Read-Only Memory (ROM), flash memory, etc.) that can store or receive software or computer code, when the software or computer code is accessed and executed by the computer, processor or hardware, the method described herein is implemented. In addition, when the general computer accesses the code for implementing the method shown herein, the execution of the code will convert the general computer into a special computer for executing the method shown herein.

[0172] Those skilled in the art can understand that the units and method steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software manner depends on the specific application and design constraints of the technical solution. The skilled person can use different methods to realize the described functions for a specific application, but such implementation should not be considered beyond the scope of the embodiments of the present application.

[0173] The above embodiments are only used to illustrate but not to limit the embodiments of the present application. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present application, and all equivalent technical solutions belong to the scope of the embodiments of the present application. The patent protection scope of the embodiments of the present application should be defined by the claims.

Claims

1. A method for resource prediction and personalized service recommendation based on large models, characterized in that, The method comprises the following steps: acquiring at least one type of historical service data; training a resource prediction model and a user analysis model using the at least one type of historical service data; acquiring at least one type of real-time service data generated by the current operation behavior of at least one user; analyzing the real-time service data using the resource prediction model to output early warning information corresponding to each user; generating a personalized service scheme corresponding to each user based on the early warning information and the real-time service data using the user analysis model; recommending the personalized service scheme corresponding to each user to the user.

2. The method of claim 1, wherein: the step of analyzing the real-time service data using the resource prediction model to output early warning information corresponding to each user comprises: for each user, performing the following steps: using the resource prediction model to predict the state of at least one target object based on real-time service data input, and obtaining a state prediction result of each target object, wherein the target object includes at least one of resource capacity, resource usage peak, and service operation condition; for each target object, when the state prediction result of the target object exceeds a preset threshold, generating early warning information based on the state prediction result and outputting the early warning information.

3. The method of claim 1, wherein: the step of generating a personalized service scheme corresponding to each user based on the early warning information and the real-time service data using the user analysis model comprises: for each user, performing the following steps: using the user analysis model to determine an early warning strategy that matches the early warning information corresponding to the user; using the user analysis model to analyze the historical operation behavior of the user and the real-time service data, and identifying the commonly used resource type and the personalized demand of the user; using the user analysis model to generate a personalized service scheme that matches the commonly used resource type and the personalized demand for the user based on the early warning strategy.

4. The method of claim 1, wherein: after the step of analyzing the real-time service data using the resource prediction model to output early warning information corresponding to each user, further comprising: pushing the early warning information to an external person, wherein the pushing mode of the early warning information includes at least one of the following: short message pushing, email pushing, voice call pushing, and platform system message pushing.

5. The method of claim 1, wherein: the step of training a resource prediction model and a user analysis model using the at least one type of historical service data comprises: performing feature extraction on each type of historical service data to obtain a feature data set; selecting a resource prediction model data set and a user analysis model data set from the feature data set; training a resource prediction model using the resource prediction model data set; training a user analysis model using the user analysis model data set.

6. The method of claim 5, wherein: the step of training a resource prediction model using the resource prediction model data set comprises: The resource prediction model dataset is divided into a resource prediction model training set and a resource prediction model test set, and the following operations are performed: Step A1: training a resource prediction model using the resource prediction model training set; Step B1: verifying whether the accuracy of the resource prediction model is higher than a preset threshold using the resource prediction model test set, and if so, performing step C1, otherwise, performing step D1; Step C1: training a user analysis model using the user analysis model dataset; Step D1: adjusting the model parameters of the resource prediction model, and performing step A1.

7. The method of claim 5, wherein the training of the user analysis model using the user analysis model dataset comprises: The user analysis model dataset is divided into a user analysis model training set and a user analysis model test set, and the following operations are performed: Step A2: training a user analysis model using the user analysis model training set; Step B2: verifying whether the accuracy of the user analysis model is higher than a preset threshold using the user analysis model test set, and if so, performing step C2, otherwise, performing step D2; Step C2: obtaining at least one type of real-time business data generated by the current operation behavior of at least one user; Step D2: adjusting the model parameters of the user analysis model, and performing step A2.

8. The method of claim 1, wherein the obtaining of at least one type of historical business data comprises: Collecting at least one type of historical data stored; Performing data processing on each type of historical data to obtain corresponding historical business data.

9. The method of any one of claims 1 to 8, wherein the obtaining of at least one type of real-time business data generated by the current operation behavior of at least one user comprises: Collecting at least one type of real-time data generated by the current operation behavior of at least one user; For each user, performing data processing on each type of real-time data collected to obtain corresponding real-time business data. comprises: A data acquisition module for acquiring at least one type of historical business data; Obtaining at least one type of real-time business data generated by the current operation behavior of at least one user; 10. An apparatus for resource prediction and personalized service recommendation based on a large model, characterized in that, A model training module for training a resource prediction model and a user analysis model using the at least one type of historical business data acquired by the data acquisition module; A model calling module for analyzing the real-time business data using the resource prediction model trained by the model training module, and outputting the corresponding early warning information for each user; Generating a personalized service plan corresponding to each user based on the early warning information and the real-time business data acquired by the data acquisition module using the user analysis model; A scheme recommendation module for recommending the personalized service plan generated by the model calling module to each user. ​ ​ ​

Images