Integrated management method and device of energy internet marketing service system

Abstract

The invention discloses an integrated management method and device for an energy internet marketing service system, and relates to the technical field of marketing management. According to the main technical scheme, the method comprises the steps of obtaining a plurality of business requirements related to an energy marketing service; respectively extracting data to be integrated from the system data of each business system according to business requirements, wherein each business system is a different data source of the energy internet marketing service system; mapping the data fields corresponding to the to-be-integrated data into the data fusion model to obtain client fusion data; and a prediction model corresponding to each business demand is constructed based on the customer fusion data, and the plurality of prediction models are integrated and stored in an energy internet marketing service system, so that business decision making can be carried out on the energy marketing service according to a prediction result. The method mainly aims at promoting intelligence of energy internet marketing service and improving the business decision-making capability corresponding to the marketing service. The method is used for integrated management of the energy internet marketing service system.

Description

Technical Field

[0001] The present application relates to the field of marketing management technology, and in particular to an integrated management method and device for an energy Internet marketing service system. Background Art

[0002] With the rapid development of energy Internet, the role of marketing service system in energy industries such as electricity is becoming increasingly significant. It not only supports the efficient operation of the energy market, but also provides customers with personalized and intelligent service experience.

[0003] At present, the existing energy Internet marketing service system is usually based on the traditional IT architecture, including but not limited to business systems such as marketing and business management, metering management, customer service, market management, smart electricity consumption, and comprehensive management. Although these business systems can realize basic business functions, different business systems operate independently, resulting in the dispersion of relevant marketing service data in multiple business systems, which not only increases the complexity of data management, but also hinders cross-departmental and cross-business line data sharing and analysis, thus affecting the intelligence of energy Internet marketing services and reducing the corresponding business decision-making capabilities of marketing services. Summary of the invention

[0004] In view of the above problems, the present application provides an integrated management method and device for an energy Internet marketing service system, the main purpose of which is to promote the intelligence of energy Internet marketing services and enhance the business decision-making capabilities corresponding to marketing services.

[0005] In order to solve the above technical problems, this application proposes the following solutions:

[0006] In a first aspect, the present application provides an integrated management method for an energy internet marketing service system, the method comprising:

[0007] Capture multiple business requirements related to energy marketing services;

[0008] Extracting the data to be integrated from the system data corresponding to each business system according to the business requirements, wherein the business system is the system that the energy internet marketing service system relies on when performing energy marketing services;

[0009] Mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data;

[0010] A prediction model corresponding to each of the business needs is constructed based on the customer fusion data, and multiple prediction models are integrated and stored in the energy Internet marketing service system so as to make business decisions on energy marketing services based on the prediction results.

[0011] In a second aspect, the present application provides an integrated management device for an energy internet marketing service system, the device comprising:

[0012] an acquisition unit, for acquiring multiple business requirements related to energy marketing services;

[0013] An extraction unit, used to extract the data to be integrated from the system data corresponding to each business system according to the business requirements, wherein the business system is the system that the energy Internet marketing service system relies on when performing energy marketing services;

[0014] A fusion unit, used for mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data;

[0015] A processing unit is used to construct a prediction model corresponding to each of the business needs based on the customer fusion data, and integrate and store multiple prediction models into the energy Internet marketing service system so as to make business decisions on energy marketing services based on the prediction results.

[0016] In order to achieve the above-mentioned purpose, according to the third aspect of the present application, a storage medium is provided, which includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the integrated management method of the energy Internet marketing service system of the above-mentioned first aspect.

[0017] In order to achieve the above-mentioned purpose, according to the fourth aspect of the present application, a processor is provided, which is used to run a program, wherein the program executes the integrated management method of the energy Internet marketing service system of the above-mentioned first aspect when running.

[0018] By means of the above-mentioned technical scheme, the present application provides an integrated management method and device for an energy Internet marketing service system. When the energy Internet marketing service system needs to be integrated and managed, first obtain multiple business requirements related to the energy marketing service, and then extract the data to be integrated from the system data corresponding to each business system according to the business requirements. The business system is the system that the energy Internet marketing service system relies on when performing energy marketing services. Then, the data fields corresponding to the data to be integrated are mapped to the data fusion model to obtain customer fusion data. Finally, a prediction model corresponding to each business requirement is constructed based on the customer fusion data, and multiple prediction models are integrated and stored in the energy Internet marketing service system, so as to make business decisions on energy marketing services based on the prediction results. Through the technical solution provided in this application, it is possible to obtain the data to be integrated related to energy marketing services in various business systems according to business needs, and use the data fusion model to obtain customer fusion data, breaking the data silos between the original business systems, realizing data integration, reducing the complexity of data management, and realizing cross-departmental and cross-business line data sharing and analysis. Then, by using customer fusion data to build a prediction model corresponding to the business needs, and integrating it into the energy Internet marketing service system, the prediction results of each prediction model can be used to support business decisions, thereby effectively promoting the intelligence of energy Internet marketing services and improving the corresponding business decision-making capabilities of marketing services.

[0019] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0021] Figure 1 A flow chart of an integrated management method of an energy Internet marketing service system provided by an embodiment of the present application is shown;

[0022] Figure 2 A flow chart of an integrated management method of another energy Internet marketing service system provided by an embodiment of the present application is shown;

[0023] Figure 3 A block diagram showing the composition of an integrated management device for an energy Internet marketing service system provided by an embodiment of the present application is shown;

[0024] Figure 4 A block diagram showing the composition of another integrated management device of an energy Internet marketing service system provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0025] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0026] At present, the existing energy Internet marketing service system is usually based on the traditional IT architecture, including but not limited to business systems such as marketing and business management, metering management, customer service, market management, smart electricity consumption, and comprehensive management. Although these business systems can realize basic business functions, different business systems operate independently, resulting in the dispersion of relevant marketing service data in multiple business systems, which not only increases the complexity of data management, but also hinders cross-departmental and cross-business line data sharing and analysis, thus affecting the intelligence of energy Internet marketing services and reducing the corresponding business decision-making capabilities of marketing services.

[0027] To this end, the embodiment of the present application provides an integrated management method for an energy internet marketing service system, through which the intelligence of energy internet marketing services can be promoted and the business decision-making ability corresponding to the marketing services can be improved. The specific execution steps are as follows: Figure 1 As shown, at least 101-104 are included.

[0028] 101. Obtain multiple business requirements related to energy marketing services.

[0029] It should be noted that in this embodiment, multiple business requirements related to energy marketing services include but are not limited to customer service, market analysis, sales forecasting, energy consumption analysis, customer behavior pattern recognition, and price strategy formulation, etc. In this step, the business requirements can be determined by conducting demand surveys by various departments or business lines, such as conference communication, questionnaire surveys, etc., or by collating historical business requirements related to energy marketing services by various departments or business lines or analyzing industry trend research, etc., which is not limited in this embodiment.

[0030] 102. Extract the data to be integrated from the system data corresponding to each business system according to business requirements.

[0031] Among them, the business system is used to characterize the system that the energy Internet marketing service system relies on in the process of performing energy marketing services.

[0032] In this step, each business system includes but is not limited to marketing and business management, metering management, customer service, market management, smart electricity consumption, comprehensive management and other major business systems. These business systems are all systems that the energy Internet marketing service system relies on when performing energy marketing services, that is, each business system is a different data source of the energy Internet marketing service system, and the energy Internet marketing service is realized through the cooperation between various business systems. Since multiple business needs have been determined, the data to be integrated is the data required to realize multiple business needs, that is, the extraction of the data to be integrated must correspond to the business needs. If sales forecasts are required, it may be necessary to extract sales records, customer information, historical electricity consumption, etc. from the system data of each business system. Specifically, a corresponding data dictionary can be created in advance for each business need, that is, the source, field name, data type, update frequency, etc. of the required data can be determined. Data directly related to marketing services, such as customer electricity consumption, service request records, transaction details and customer feedback, are extracted from the system data of these systems through the data dictionary. The data extraction process can use ETL (Extract, Transform, Load) tools to ensure the quality and consistency of the data.

[0033] It should be noted that after obtaining the data to be integrated, in order to ensure the integrity and accuracy of the data to be integrated, the data to be integrated can be preprocessed, including removing duplicates, filling missing values, correcting erroneous data, etc.

[0034] 103. Map the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data.

[0035] In this step, the data fusion model is used to integrate and unify heterogeneous data from different business systems to create a comprehensive and consistent customer view. By mapping the data fields corresponding to the data to be integrated into the data fusion model, customer fusion data can be obtained, and the customer fusion data refers to the customer-related information set from multiple business systems that is integrated and unified through the data fusion model, that is, the customer view. The customer view contains all customer-related information, such as personal information, transaction history, service requests, feedback, etc., that is, information about various dimensions of the customer. Specifically, the customer field in the data field that represents the customer can be used as the customer entity of the data fusion model, such as user number, device ID, etc., and the remaining fields can be mapped to the entity attributes of the corresponding customer entity, so as to obtain the entity data corresponding to each customer entity, that is, customer fusion data. By fusing the data to be integrated extracted from different business systems with customers as the main body, it can ensure that the final customer fusion data accurately reflects the real situation of the customer, thereby providing a reliable foundation for the subsequent prediction model.

[0036] 104. Build a prediction model corresponding to each business demand based on customer fusion data, and integrate and store multiple prediction models into the energy Internet marketing service system so as to make business decisions on energy marketing services based on the prediction results of the prediction model.

[0037] In this step, since the customer fusion data related to energy marketing services in various business systems has been integrated, in order to better meet business needs, it is possible to provide assistance for business decisions of energy marketing services and make them more intelligent. Specifically, a prediction model corresponding to each business demand can be constructed based on customer fusion data. The construction of the prediction model can use machine learning algorithms such as regression analysis, decision trees, neural networks, etc. Among them, one business demand corresponds to a different prediction model. The prediction model includes but is not limited to a customer churn prediction model, a sales prediction model, or a price sensitivity model. Integrating multiple constructed prediction models into the energy Internet marketing service system allows business decision makers to make more informed decisions based on the prediction results, such as adjusting marketing strategies, optimizing customer service processes, or improving product pricing. The integrated storage process can encapsulate each prediction model into an independent executable, i.e., a model service. The model service can be used as a functional module or microservice, and a clear input and output interface is defined for the executable, so that the executable can be called as an independent component. An API (application programming interface) can also be developed to expose the functions of the executable. The prediction model can also be deployed as an executable using container technology (such as Docker) and container orchestration tools (such as Kubernetes), usually on a cloud platform or a local server. This embodiment does not limit this.

[0038] For example, it is assumed that the business demand is to predict the customer churn rate in the next month so as to take measures to reduce customer churn and improve customer satisfaction and loyalty. First, according to the business demand, it can be known that the corresponding prediction target is to predict the customer churn rate in the future period of time. Then, for this prediction target, according to the customer fusion data obtained in step 103, including basic customer information, historical transaction records and service request records, etc., select features related to customer churn, such as customer satisfaction score, the last transaction time, service request frequency and the number of historical complaints, etc. Then, select random forest as the prediction model to be trained, and divide the obtained feature data into a training set and a test set, use the training set to train the prediction model, and use the test set to evaluate the performance of the model to obtain the prediction model corresponding to the business demand, that is, the trained customer churn rate prediction model. Finally, the trained customer churn rate prediction model is deployed to the energy Internet marketing service system, and the newly collected customer fusion data is predicted in real time by calling the customer churn rate prediction model to obtain the customer churn rate in the next month, that is, the prediction result.

[0039] The specific business decision-making process for this prediction result is as follows: customers are divided into different risk levels according to the prediction results, and personalized retention plans are formulated for customers of different risk levels. For example, for high-risk customers, coupons and upgraded services can be provided; for medium-risk customers, customer satisfaction can be improved by increasing customer interaction and service quality; and for low-risk customers, long-term relationship building can be focused on, such as launching membership plans.

[0040] In addition, in order to ensure that the prediction model can better respond to the ever-changing market environment and customer needs, the multiple prediction models can be retrained regularly, such as once a week or once a month, or the performance of the multiple prediction models can be monitored in real time, and whether to retrain the multiple prediction models can be determined based on the monitoring results.

[0041] Based on the above Figure 1 It can be seen from the implementation method that the integrated management method of an energy Internet marketing service system provided by the present application is to first obtain multiple business needs related to the energy marketing service when the energy Internet marketing service system needs to be integrated and managed, and then extract the data to be integrated from the system data of each business system according to the business needs. Each business system is a different data source for the energy Internet marketing service system, and then map the data fields corresponding to the data to be integrated to the data fusion model to obtain customer fusion data, and finally build a prediction model corresponding to each business need based on the customer fusion data, and integrate and store multiple prediction models in the energy Internet marketing service system, so as to make business decisions on energy marketing services based on the prediction results. Through the technical solution provided in this application, it is possible to obtain the data to be integrated related to energy marketing services in various business systems according to business needs, and use the data fusion model to obtain customer fusion data, breaking the data silos between the original business systems, realizing data integration, reducing the complexity of data management, and realizing cross-departmental and cross-business line data sharing and analysis. Then, by using customer fusion data to build a prediction model corresponding to the business needs, and integrating it into the energy Internet marketing service system, the prediction results of each prediction model can be used to support business decisions, thereby effectively promoting the intelligence of energy Internet marketing services and improving the corresponding business decision-making capabilities of marketing services.

[0042] Further, the preferred embodiment of the present application is in the above Figure 1 Based on this, a detailed description of the integrated management process of the energy Internet marketing service system is given. The specific steps are as follows: Figure 2 As shown, including 201-209:

[0043] 201. Obtain multiple business requirements related to energy marketing services.

[0044] This step is combined with the description of step 101 in the above method, and the same contents will not be repeated here.

[0045] 202. According to business requirements, extract the data to be integrated from the system data corresponding to each business system.

[0046] This step is combined with the description of step 102 in the above method, and the same content will not be repeated here. It should be noted that the specific execution process of extracting the data to be integrated from the system data corresponding to each business system according to business needs is: obtaining a data dictionary pre-built for each business need; performing field matching in the system data corresponding to each business system according to the data dictionary to obtain the fields to be integrated; determining the fields to be extracted from the fields to be integrated using preset screening rules; extracting the fields to be extracted to obtain the data to be integrated.

[0047] In this step, the data dictionary can determine in advance what data is needed to support the realization of these business requirements based on the details of the business requirements. That is, based on business requirements, determine which data fields are required. For example, for customer service needs, data fields such as customer ID, contact information, and service history may be required. Collect metadata for each data field, including field name, field type, and field description. Use tables, databases, or specialized data dictionary management systems to record the above information to ensure that the data dictionary is easy to update and maintain. Once the business requirements are determined, the corresponding data dictionary can be obtained from the above-mentioned tables, databases, or specialized data dictionary management systems.

[0048] With the guidance of the data dictionary, the fields to be integrated that meet the business needs can be accurately matched from each business system. Since each business system is a different data source of the energy Internet marketing system, it is possible to determine from each business system which business systems contain the data fields listed in the data dictionary. The fields to be integrated obtained in the matching process need to be consistent in type and semantically similar to the fields in the data dictionary. Therefore, a screening rule can be set, which is used to characterize the rules for screening based on the type and semantics of the field. Specifically, the fields in the data dictionary can be compared with the types of the fields to be integrated obtained by matching the fields in the business system. If they are consistent, the semantic similarity between the two is calculated. When the semantic similarity is higher than the preset similarity threshold, it is determined that the field to be integrated needs to be extracted. Otherwise, the field to be integrated is removed, thereby ensuring that the final extracted field is accurate, that is, the data to be integrated is accurate, thereby improving the accuracy of subsequent data fusion. In the extraction process, the fields to be extracted can be extracted by writing SQL query statements or other data extraction tools to obtain the data to be integrated.

[0049] 203. Map the target field in the data field to the customer entity of the data fusion model.

[0050] The target field is the field representing the customer.

[0051] In this step, the target field can be a customer ID or a similar field that can uniquely identify each customer, such as a user number, device ID, etc., such as a smart meter ID. Data cleaning is performed on the target field to handle missing values ​​or inconsistent records. At the same time, if the target field has different representations in different business systems, a specified format can be pre-set and converted according to the specified format to obtain a unified format. The target field is mapped to the customer entity in the data fusion model to facilitate linking and aggregation of all information related to a specific customer.

[0052] 204. Map the remaining fields in the data field to entity attributes corresponding to the customer entity to obtain entity data corresponding to each customer entity.

[0053] In this step, the customer entity and its attributes can be organized in advance into a structured data format, such as JSON, XML, or a relational database table, and all remaining customer-related data fields except the target field can be mapped one by one according to the attributes of the customer entity, thereby merging the same attribute data from different business systems to construct detailed information for each customer entity, namely, entity data.

[0054] 205. Entity data corresponding to all customer entities are used as customer fusion data.

[0055] In this step, all customer entities and their attribute data are collected together to form a complete customer fusion data set, namely, customer fusion data. In order to access and further analyze the customer fusion data, the customer fusion data can be stored in a centralized location, such as a data warehouse or data lake of the energy Internet marketing system, as a basis for subsequent analysis and prediction.

[0056] 206. Determine the corresponding forecast target based on each business need.

[0057] In this step, since each business requirement has a clear business goal, the qualitative business goal is converted into a quantifiable forecasting goal, such as customer churn rate, sales, demand, etc. Specifically, one or more measurable indicator data and corresponding time scales can be defined in advance for each business requirement, where the indicator data can be a ratio (such as customer churn rate, inventory turnover rate), frequency (such as customer repurchase frequency), amount (such as total sales, average order value), quantity (such as demand, click volume) or other numerical data, and the time scale can be a short-term scale (such as one day, one week, one month) or a long-term scale (such as one quarter, half a year, one year), etc. There is a corresponding relationship between the indicator data and the time scale to ensure that each indicator data has a clear time scale corresponding to it, so as to clearly define the forecasting goal. For example, "predict the customer churn rate of the next quarter" or "predict the total sales of the next year", etc. When a business demand corresponds to an indicator data, the corresponding prediction target can be determined based on the indicator data and the time scale corresponding to the indicator data. When the business demand corresponds to multiple indicator data, the degree of fit between each indicator data and the business demand can be calculated separately, and the corresponding prediction target can be determined based on the indicator data with the highest fit and its corresponding time scale. The fit can use statistical methods (such as correlation analysis, principal component analysis) to evaluate the relationship between different indicator data, identify which indicators have strong correlations, and which indicators provide unique information, so as to determine the indicator data with the highest fit. In addition, when the business demand corresponds to multiple indicator data, multiple prediction targets can be directly obtained based on multiple indicator data and the time scale corresponding to each indicator data, and the multiple prediction targets can be taken as a whole to train a multi-objective prediction model in step 207.

[0058] Example 1:

[0059] Assuming that the business demand is to improve customer satisfaction, the corresponding indicator data is one, namely the customer net promoter score (NPS), and the time scale corresponding to this indicator data is one quarter, then the corresponding prediction target is: predict the customer net promoter score for the next quarter. Among them, the business demand is to improve customer satisfaction, and the customer net promoter score (NPS) is a widely used indicator to measure the degree to which customers are willing to recommend a company's products or services to others. Since NPS is a ratio indicator and changes in customer satisfaction often take a certain amount of time to manifest, it is reasonable to set one quarter as the time scale in advance.

[0060] Example 2:

[0061] Assume that the business demand is to optimize the power supply to improve the user's power experience and energy efficiency. The corresponding indicator data are: user power consumption prediction accuracy, peak power consumption prediction accuracy and energy efficiency improvement rate. The corresponding time scales are: one month, one quarter and one year. In order to determine the indicator data that best fits the business needs, correlation analysis can be used to evaluate the strength of the relationship between each indicator data and the business needs. The strength of the relationship can be represented by the average value of the Pearson correlation coefficient (applicable to linear relationships) and the Spearman rank correlation coefficient (applicable to non-linear relationships). If, after calculation, the indicator data with the greatest relationship strength is the energy efficiency improvement rate, the final prediction goal is to predict the energy efficiency improvement rate for the next year.

[0062] 207. Extract corresponding feature data sets from the customer fusion data according to each prediction target, and train multiple prediction models based on the multiple feature data sets.

[0063] In this step, features that may affect the prediction target can be selected based on business knowledge and data exploration. It is also possible to calculate the correlation coefficient between each feature and the prediction target, such as the Pearson correlation coefficient or the Spearman rank correlation coefficient, and to delineate the feature data related to the prediction target by setting a correlation coefficient threshold, which is not limited in this embodiment. Different prediction models are trained using different feature data sets, and the multiple prediction models can use a variety of machine learning algorithms, such as linear regression, random forest, neural network, etc. Specifically, the prediction model adapted for each prediction target can be determined based on the target characteristics of the prediction target or the data characteristics of the feature data set. Among them, the target characteristics include target type and target complexity, the target type includes classification type and regression type, and the target complexity is used to characterize whether the feature data set is in a linear relationship with the prediction target. Data characteristics include data scale, data type and data quality. The data scale is used to characterize the data volume of the feature data set, the data type includes numerical type and categorical type, and the data quality includes whether there are missing values ​​or outliers in the feature data set. The target type, target complexity, data scale, data type and data quality can be used as decision indicators, and a correlation table between the decision indicators and the prediction model can be pre-defined. Based on the correlation table, the prediction model categories corresponding to the target type, target complexity, data scale, data type and data quality can be determined respectively. The prediction model with the most occurrences can be selected as the final prediction model of the prediction target. Different weights can also be set for the above-mentioned decision indicators, and the prediction model with the largest score obtained by weighted calculation can be used as the final prediction model of the prediction target. This embodiment does not limit this. Through such a prediction model selection method, it can be ensured that each prediction target can select its adapted prediction model for training, thereby ensuring the accuracy of subsequent business decisions. In order to ensure the accuracy of the extracted feature data set, the accuracy of the prediction model is ensured. In view of the above description of the feature data set corresponding to the correlation coefficient, the specific execution process of extracting the corresponding feature data set in the customer fusion data according to each prediction target is as follows: determine the correlation coefficient type corresponding to each feature data according to the data performance of each feature data in the customer fusion data within a specified period, and the data performance includes the size of linear deviation and the frequency of occurrence of outliers; calculate the correlation coefficient between each feature data and each prediction target according to the correlation coefficient type; use the feature data with a correlation coefficient higher than a preset correlation coefficient threshold as the feature data set for each prediction target, and one correlation coefficient threshold corresponds to one prediction target.

[0064] Among them, the specified period can be set according to the prediction target, such as short-term scale (such as one day, one week, one month) or long-term scale (such as one quarter, half a year, one year), etc. The selection of the correlation coefficient type depends on the data performance of each feature data in the specified period, specifically the linear deviation size and the frequency of occurrence of outliers for each feature data. The linear deviation size refers to the degree of deviation of the data points of the feature data in the specified period relative to the predicted value of the linear regression model. Specifically, the residual of the linear regression model (that is, the difference between the actual value and the predicted value) can be calculated to evaluate the linear deviation size. The smaller the standard deviation of the residual, the smaller the linear deviation. By comparing the residual calculated for each feature data, that is, the linear deviation size, with the preset deviation threshold, the distribution of the feature data in the specified period can be determined. The frequency of occurrence of outliers refers to the proportion of outliers in the feature data in the specified period. Outliers refer to values ​​that deviate significantly from other observed values. Specifically, the interquartile range or Z-score (standardized score) can be used to identify outliers. By calculating the frequency of occurrence of outliers in the specified period and comparing it with the preset frequency threshold, the occurrence of outliers in the feature data in the specified period can be determined. By combining the above distribution and the occurrence of outliers, the data performance of each feature data can be determined. That is, when the linear deviation is less than or equal to the preset residual threshold and the frequency of occurrence of outliers in a specified period is less than or equal to the preset frequency threshold, the feature data is a linear correlation coefficient type, and the calculated Pearson correlation coefficient is used as the correlation coefficient of the feature data. When the linear deviation is greater than the preset residual threshold and / or the frequency of occurrence of outliers in a specified period is greater than the preset frequency threshold, the feature data is a nonlinear correlation coefficient type, and the calculated Spearman rank correlation coefficient is used as the correlation coefficient of the feature data.

[0065] The correlation coefficient threshold can determine which feature data may have an impact on the prediction target through domain knowledge and prior information, and use the lowest correlation coefficient among these feature data as the correlation coefficient threshold. In this way, a corresponding correlation coefficient threshold can be set for each prediction target. In addition, the correlation coefficient threshold can also be adjusted according to the accuracy requirements of the prediction model.

[0066] For the above description of outlier identification, for example,

[0067] Use the interquartile range to identify outliers:

[0068] formula:

[0069] IQR=Q3-Q1

[0070] Lower limit = Q1-1.5×IQR

[0071] Upper limit=Q3+1.5×IQR

[0072] Among them, Q1 (first quartile): the maximum value of the smallest 25% of the data in the data set; Q3 (third quartile): the minimum value of the largest 25% of the data in the data set; IQR (interquartile range): the distance between Q3 and Q1, that is, the range of the middle 50% of the data. Any value lower than Q1-1.5×IQR is considered a lower limit outlier; any value higher than Q3+1.5×IQR is considered an upper limit outlier.

[0073] Use the Z-score method to identify outliers:

[0074] Formula: Z = σx-μ

[0075] Where x is the value of the data point; μ is the mean of the data set; and σ is the standard deviation of the data set. Typically, any data point with a Z-score greater than 3 or less than -3 is considered an outlier.

[0076] 208. Encapsulate multiple prediction models into multiple executable bodies according to corresponding business requirements.

[0077] The executable is used to represent the model service that defines the effective duration and calling method.

[0078] Use Docker containers and other methods to package each prediction model and its dependencies into multiple executables, so that they become a function or microservice that can be called repeatedly, namely, a model service. Define a clear calling method and effective duration for each model service. The calling method includes API interface, parameter description, authentication and authorization, calling frequency and limitation, and error handling, etc. The effective duration refers to the available duration of each model service obtained after encapsulating the prediction model. By setting the effective duration and calling method, the multiple executables that encapsulate the prediction model can be effectively called within their corresponding available duration. Therefore, when receiving the corresponding business needs related to energy marketing services, the business needs can be predicted by calling the prediction model, so as to provide assistance for business decisions of energy marketing services based on the prediction results.

[0079] 209. Integrate and store multiple executables into the energy internet marketing service system to facilitate calling in the energy internet marketing service system.

[0080] In this step, a suitable deployment environment can be selected in the energy internet marketing service system based on whether the computing resources and network bandwidth are sufficient, such as the cloud server, local server or edge device corresponding to the energy internet marketing service system, etc., which is not limited in this embodiment. By deploying the model service to the energy internet marketing service system, it becomes a part of the energy internet marketing service system, so that it can be called in real time or in batches. Therefore, when receiving the corresponding business demand related to the energy marketing service, the business demand can be predicted by calling the prediction model, so as to provide assistance to the business decision of the energy marketing service according to the prediction results.

[0081] Furthermore, as can be seen from the aforementioned step 208, the executable body also involves the effective duration. Therefore, in order to ensure that the prediction model can better cope with the ever-changing market environment and customer needs, thereby ensuring the accuracy of the prediction results. The method also includes: when the remaining duration of the effective duration is less than the preset duration threshold, a new prediction model corresponding to each business demand is rebuilt according to the latest customer fusion data; multiple new prediction models are integrated and stored in the energy Internet marketing service system, and the original multiple prediction models are deleted according to the call status.

[0082] Among them, since the prediction model will set the effective duration when the construction is completed, the remaining duration of the effective duration can be monitored by comparing with the current moment. In order to enable the update of the prediction model to obtain the corresponding new prediction model before the end time corresponding to the effective duration, a duration threshold corresponding to the effective duration can be preset. The setting of the duration threshold can be based on the timeliness of business needs, the frequency of data updates, or the decay rate of the model prediction performance. This embodiment does not limit this. Once it is monitored that the remaining effective duration of a prediction model is lower than the preset duration threshold, the system automatically triggers the reconstruction process of the prediction model, that is, repeating the aforementioned steps 201-208, and reconstructing the new prediction model corresponding to each business demand according to the latest customer fusion data, and integrating and storing multiple new prediction models into the energy Internet marketing service system.

[0083] In order to avoid the coexistence of two prediction models with duplicate functions, after integrating and storing multiple new prediction models into the energy Internet marketing service system, the original prediction models in the system need to be deleted. However, if the deletion operation occurs when the original prediction model is being called, it may interrupt the customer's prediction operation and thus reduce the user experience. Therefore, in order to enhance the user experience, multiple prediction models can be deleted in sequence according to the calling status. The specific execution process is: if the calling status of the prediction model is not called, the prediction model is deleted; if the calling status of the prediction model is called, the prediction model is deleted when the calling status of the prediction model changes to not called.

[0084] Among them, the call status includes being called and not being called, which can be determined by querying the current status of the model service corresponding to the encapsulated prediction model or the call counter. If the call status is not being called, the prediction model deletion operation is performed, that is, the prediction model is uninstalled from the energy Internet marketing service system to release the resources it occupies. If the call status is called, a monitoring mechanism is set up. When the prediction model is no longer called, that is, when its call status changes to not being called, the deletion operation is immediately performed.

[0085] Furthermore, as a response to the above Figure 1-2 The implementation of the method embodiment shown in the figure, the embodiment of the present application provides an integrated management device for an energy internet marketing service system, which is used to promote the intelligence of energy internet marketing services and enhance the business decision-making capabilities corresponding to marketing services. The embodiment of the device corresponds to the aforementioned method embodiment. For ease of reading, this embodiment will no longer repeat the details of the aforementioned method embodiment one by one, but it should be clear that the device in this embodiment can correspond to all the contents of the aforementioned method embodiment. Specifically, Figure 3 As shown, the device comprises:

[0086] An acquisition unit 31, used to acquire multiple business requirements related to energy marketing services;

[0087] An extraction unit 32, used to extract the data to be integrated from the system data corresponding to each business system according to the business requirements, each of the business systems being a different data source of the energy internet marketing service system;

[0088] A fusion unit 33, used for mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data;

[0089] The processing unit 34 is used to construct a prediction model corresponding to each of the business needs based on the customer fusion data, and integrate and store multiple prediction models into the energy Internet marketing service system, so as to make business decisions on energy marketing services based on the prediction results of the prediction models.

[0090] Further, such as Figure 4 As shown, the extraction unit 32 includes:

[0091] An acquisition module 321 is used to acquire a pre-built data dictionary for each of the business requirements;

[0092] A screening module 322, which uses a preset screening rule to determine the fields to be extracted from the fields to be integrated, wherein the screening rule is used to represent a rule for screening based on the type and semantics of the field;

[0093] The extraction module 323 extracts the fields to be extracted to obtain the data to be integrated.

[0094] Further, such as Figure 4 As shown, the fusion unit 33 includes:

[0095] A first determination module 331 is used to map a target field in the data field to a customer entity of the data fusion model, wherein the target field is a customer field;

[0096] A mapping module 332 is used to map the remaining fields in the data field to entity attributes corresponding to the customer entity to obtain entity data corresponding to each of the customer entities;

[0097] The second determining module 333 is used to use the entity data corresponding to all the customer entities as the customer fusion data.

[0098] Further, such as Figure 4 As shown, the processing unit 34 includes:

[0099] A third determination module 341 is used to determine a corresponding prediction target according to each of the business requirements;

[0100] An extraction module 342, configured to extract a corresponding feature data set from the customer fusion data according to each prediction target, and train a plurality of prediction models based on the plurality of feature data sets;

[0101] The encapsulation module 343 is used to encapsulate the plurality of prediction models into a plurality of executable bodies according to the corresponding business requirements. The executable bodies are used to represent the model services that define the effective duration and calling method;

[0102] The integration module 344 is used to integrate and store the plurality of executable bodies into the energy internet marketing service system so as to facilitate calling in the energy internet marketing service system.

[0103] Further, such as Figure 4 As shown, the extraction module 432 is specifically used for:

[0104] Determine the type of correlation coefficient corresponding to each feature data according to the data performance of each feature data in the customer fusion data within a specified period, wherein the data performance includes the linear deviation size and the frequency of occurrence of abnormal values;

[0105] Calculate the correlation coefficient between each of the feature data and each of the prediction targets according to the correlation coefficient type;

[0106] The feature data whose correlation coefficient is higher than a preset correlation coefficient threshold is used as a feature data set of each prediction target, and one correlation coefficient threshold corresponds to one prediction target.

[0107] Further, such as Figure 4 As shown, the device also includes:

[0108] An updating module 345 is configured to reconstruct a new prediction model corresponding to each of the business requirements according to the latest customer fusion data when the remaining duration of the effective call duration is less than a preset duration threshold;

[0109] The integration module 343 is also used to integrate and store the multiple new prediction models into the energy Internet marketing service system, and delete the original multiple prediction models according to the call status.

[0110] Further, such as Figure 4 As shown, the update module 345 is specifically used for:

[0111] If the calling state of the prediction model is not called, deleting the prediction model;

[0112] If the calling state of the prediction model is called, the prediction model is deleted when the calling state of the prediction model is changed to not called.

[0113] Furthermore, an embodiment of the present application further provides a storage medium, wherein the storage medium is used to store a computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the above Figure 1-2 The integrated management method of the energy Internet marketing service system described in.

[0114] Furthermore, the embodiment of the present application also provides a processor, the processor is used to run a program, wherein the program executes the above Figure 1-2 The integrated management method of the energy Internet marketing service system described in.

[0115] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0116] It is understandable that the related features in the above methods and devices can be referenced to each other. In addition, the "first", "second" and the like in the above embodiments are used to distinguish the embodiments, but do not represent the advantages and disadvantages of the embodiments.

[0117] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0118] The algorithm and display provided herein are not inherently related to any particular computer, virtual system or other device. Various general purpose systems can also be used together with the teachings based on this. According to the above description, it is obvious that the structure required for constructing such systems. In addition, the application is not directed to any specific programming language either. It should be understood that various programming languages ​​can be utilized to realize the content of the application described herein, and the description of the specific language above is for the purpose of disclosing the best mode of implementation of the application.

[0119] In addition, the memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0120] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0121] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0122] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0123] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0124] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0125] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0126] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include transitory media such as modulated data signals and carrier waves.

[0127] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0128] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0129] The above are only embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included within the scope of the claims of the present application.

Claims

1. An integrated management method for an energy internet marketing service system, characterized in that: The method comprises: Capture multiple business requirements related to energy marketing services; Extracting the data to be integrated from the system data corresponding to each business system according to the business requirements, wherein the business system is the system that the energy internet marketing service system relies on when performing energy marketing services; Mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data; A prediction model corresponding to each business demand is constructed based on the customer fusion data, and multiple prediction models are integrated and stored in the energy Internet marketing service system, so as to make business decisions on energy marketing services based on the prediction results of multiple prediction models.

2. The method according to claim 1, characterized in that According to the business requirements, the data to be integrated related to the energy marketing service are extracted from the system data corresponding to each business system, including: Obtain a pre-built data dictionary for each of the business requirements; perform field matching in the system data corresponding to each business system according to the data dictionary to obtain the fields to be integrated; Determining the fields to be extracted from the fields to be integrated by using preset screening rules, wherein the screening rules are used to represent the rules for screening based on the type and semantics of the fields; The fields to be extracted are extracted to obtain the data to be integrated.

3. The method according to claim 1, characterized in that: Mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data, including: Mapping a target field in the data field to a customer entity of the data fusion model, wherein the target field is a customer field; Mapping the remaining fields in the data field except the target field to the entity attributes corresponding to the customer entity to obtain the entity data corresponding to each of the customer entities; The entity data corresponding to all the customer entities are used as the customer fusion data.

4. The method according to claim 1, characterized in that Constructing a prediction model corresponding to each of the business needs based on the customer fusion data, and integrating and storing multiple prediction models in the energy Internet marketing service system, including: Determine corresponding forecast targets according to each of the business requirements; Extracting a corresponding feature data set from the customer fusion data according to each prediction target, and training a plurality of prediction models based on a plurality of the feature data sets; Encapsulating the plurality of prediction models into a plurality of executable bodies according to the corresponding business requirements, wherein the executable bodies are used to represent the model services that define the effective duration and calling method; The plurality of executable bodies are integrated and stored in the energy internet marketing service system so as to be called in the energy internet marketing service system.

5. The method according to claim 4, characterized in that Extracting a corresponding feature data set from the customer fusion data according to each prediction target includes: determining a correlation coefficient type corresponding to each feature data according to a data performance of each feature data in the customer fusion data within a specified period, wherein the data performance includes a linear deviation size and an abnormal value occurrence frequency; Calculate the correlation coefficient between each feature data and each prediction target according to the correlation coefficient type; The feature data whose correlation coefficient is higher than a preset correlation coefficient threshold is used as a feature data set of each prediction target, and one correlation coefficient threshold corresponds to one prediction target.

6. The method according to claim 4, characterized in that The method further comprises: When the remaining time of the effective time is less than the preset time threshold, a new prediction model corresponding to each business demand is rebuilt according to the latest customer fusion data; The multiple new prediction models are integrated and stored in the energy internet marketing service system, and the original multiple prediction models are deleted according to the call status.

7. The method according to claim 6, characterized in that Deleting the original multiple prediction models according to the calling status includes: If the calling status of the original prediction model is not called, deleting the original prediction model; If the calling state of the original prediction model is called, the original prediction model is deleted when the calling state of the original prediction model is changed to not called.

8. An integrated management device for an energy internet marketing service system, characterized in that: The device comprises: an acquisition unit, for acquiring multiple business requirements related to energy marketing services; An extraction unit, used to extract the data to be integrated from the system data corresponding to each business system according to the business requirements, wherein the business system is the system that the energy Internet marketing service system relies on when performing energy marketing services; A fusion unit, used for mapping the data fields corresponding to the data to be integrated into the data fusion model to obtain customer fusion data; A processing unit is used to construct a prediction model corresponding to each of the business needs based on the customer fusion data, and integrate and store multiple prediction models into the energy Internet marketing service system so as to make business decisions on energy marketing services based on the prediction results.

9. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the integrated management method of the energy Internet marketing service system as described in any one of claims 1 to 7.

10. A processor, characterized in that: The processor is used to run a program, wherein the program, when running, executes the integrated management method of the energy Internet marketing service system as described in any one of claims 1 to claim 7.

Images