An enterprise user-oriented power service self-service intelligent recommendation method

Abstract

The present application relates to the technical field of data processing, more particularly, the present application relates to a kind of enterprise user-oriented electric power service self-service intelligent recommendation method, comprising: obtaining the static feature vector of each enterprise user and each historical dynamic feature vector when accessing each electric power service;Determine the historical dynamic feature concentration of each enterprise user;Determine the user concentration of each electric power service;Determine the recommendation degree of each electric power service for each enterprise user;Based on the recommendation degree, to realize the intelligent recommendation of enterprise user-oriented electric power service self-service.The present application accurately measures the stability and diversity of user demand through double feature construction and clustering analysis;Realize business differentiation modeling, avoid the disadvantages of unified strategy;Through comprehensive evaluation, realize the dynamic adjustment of recommendation degree, adapt to preference changes, provide cold start compensation for new users, improve recommendation accuracy and personalization.

Description

Technical Field

[0001] This invention relates to the field of data processing technology. More specifically, this invention relates to an intelligent recommendation method for self-service electricity business for enterprise users. Background Technology

[0002] With the continuous development of information technology, more and more enterprise users are choosing to handle electricity-related business online. This enterprise user group is highly diverse, and their needs are also multifaceted. For a single transaction, an enterprise user typically only handles one or a few services. If they are forced to manually select the required electricity services from a vast array of options, it easily leads to low operational efficiency and a poor user experience. Therefore, providing precise service recommendations to enterprise users is crucial for optimizing the user experience.

[0003] Chinese patent document CN113469731B discloses an intelligent recommendation method, platform, and server based on big data services of the power Internet of Things. The application document obtains user data carried in service requests; determines service data matching the service request based on the user data; and recommends the determined service data to the target user. The method achieves the purpose of pushing corresponding service information to users based on consumption preference information.

[0004] However, the aforementioned patent documents rely solely on consumer preference information for user service recommendations, failing to deeply mine and analyze the dynamic behavioral characteristics generated by repeated user visits, making it difficult to measure the stability and diversity of user needs. Furthermore, different electricity services exhibit significant differences in their processing; some services are frequently handled by many companies (e.g., electricity bill payment, high concentration), while others are only occasionally handled by a few companies (e.g., power equipment upgrade applications, low concentration). The lack of differentiated modeling for different electricity services results in all services using the same recommendation strategy. For new users making their first visit, there is also a lack of an effective cold-start compensation mechanism. In addition, the calculation of recommendation scores relies solely on simple linear calculations, lacking nonlinear adjustment methods to dynamically amplify or smooth user preferences based on their stability, making it difficult to adapt to dynamic changes in user preferences. Summary of the Invention

[0005] To address the shortcomings of existing technologies that rely solely on consumer preference information for user service recommendations, failing to deeply mine and analyze the dynamic behavioral characteristics generated by repeated user visits, and thus struggling to measure the stability and diversity of user needs; furthermore, the lack of differentiated modeling for different power services leads to the application of the same recommendation strategy across all services; and the absence of an effective cold-start compensation mechanism for new users on their first visit; and the fact that recommendation degree calculation relies solely on simple linear calculations, making it difficult to adapt to dynamic changes in user preferences, this invention proposes an intelligent recommendation method for power service self-service for enterprise users. This method includes the following steps:

[0006] Any enterprise user is designated as a target user. Based on the target user's static characteristics at registration, a static feature vector is determined. Based on each historical dynamic feature of the target user when accessing various power services, a historical dynamic feature vector is determined. All historical dynamic feature vectors of the target user when accessing all power services are clustered to obtain multiple dynamic feature clusters. The historical dynamic feature concentration of the target user is determined based on the maximum and minimum distances between the cluster centers of all dynamic feature clusters. Any power service is designated as a target service, and enterprise users who have handled target services are designated as service users. The user concentration of the target service is determined based on the mean of the historical dynamic feature concentration of all service users, the mean of the cosine similarity between the static feature vectors of all service users, and the proportion of times each service user handles the target service out of the total number of times they handle all power services. The recommendation degree of the target service for the target user is determined based on the number of times the target user handles the target service, the historical dynamic feature vector concentration, and the user concentration. Based on the recommendation degree, intelligent recommendation of power service self-service for enterprise users is achieved.

[0007] The beneficial effects are as follows: By constructing dual features of static feature vectors and historical dynamic feature vectors, and using cluster analysis to quantify the concentration of user behavior, an accurate measurement of the stability and diversity of user needs is achieved, overcoming the limitations of traditional methods that rely solely on consumer preferences; by calculating the average historical dynamic feature concentration of the business user group, the average cosine similarity of static feature vectors, and the proportion of business transaction frequency, a user concentration index for the target business is constructed, enabling differentiated modeling of different power businesses and avoiding the drawbacks of a uniform recommendation strategy; by introducing a comprehensive evaluation of historical dynamic feature concentration and user concentration, nonlinear dynamic adjustment of the recommendation degree is achieved, which can dynamically amplify or smooth the user preference stability, effectively adapting to the dynamic changes in user preferences; through group similarity analysis of static feature vectors and statistical features of business transactions, an effective cold start compensation mechanism is provided for new users, and reasonable recommendations are made using the behavioral patterns of similar users, significantly improving the initial service experience of new users; the complete recommendation framework built based on dynamic feature clustering, user concentration analysis, and nonlinear recommendation degree calculation realizes intelligent recommendation of enterprise users' power business self-service, improving the accuracy and personalization of recommendations.

[0008] Furthermore, the static characteristics include enterprise size, electricity consumption type, industry classification, region, and remaining electricity.

[0009] Furthermore, the dynamic characteristics include dwell time, keyword input, and access time.

[0010] Furthermore, determining the static feature vector of the target user includes: combining the normalized result of the continuous data in the static features of the target user at the time of registration with the encoding result of the category data in the static features of the target user at the time of registration to obtain the static feature vector of the target user.

[0011] Furthermore, determining each historical dynamic feature vector of the target user when accessing each power service includes: combining the normalized result of the continuous data in each historical dynamic feature of the target user when accessing each power service with the encoding result of the categorical data in each historical dynamic feature of the target user when accessing each power service to obtain each historical dynamic feature vector of the target user when accessing each power service.

[0012] Furthermore, the clustering adopts the mean-shift clustering algorithm.

[0013] Furthermore, the concentration of the historical dynamic features satisfies:

[0014] In the formula, The concentration of historical dynamic characteristics of the target users. It is the maximum distance among the cluster centers of all dynamic feature clusters of the target user. It is the minimum distance between the cluster centers of all dynamic feature clusters of the target user. for Type curve function.

[0015] The beneficial effect is that by calculating the difference between the maximum and minimum distances between the cluster centers of all dynamic feature clusters of the target user, and using... The function performs mapping, achieving standardized quantification of the concentration of historical dynamic features; function The curve-like characteristics limit the concentration range to within Within the specified range, this facilitates subsequent calculations and comparisons.

[0016] Furthermore, the user concentration satisfies:

[0017] In the formula, For the first User concentration of individual power businesses For all those who have processed the first The mean of the historical dynamic characteristics concentration of enterprise users in the power business. For all those who have processed the first The mean cosine similarity among the static feature vectors of enterprise users in the power business. For all those who have processed the first The number of enterprise users in the electricity business For all those who have processed the first The first among the enterprise users of the power business The first enterprise user to process the The number of times a power business is conducted. For all those who have processed the first The first among the enterprise users of the power business The number of times an enterprise user handles all electricity-related services.

[0018] The beneficial effects are as follows: By comprehensively considering three dimensions—the average concentration of historical dynamic characteristics of business users, the average similarity of static characteristics, and the proportion of business processing frequency—a comprehensive evaluation model for the concentration of electricity business users is constructed. This model can accurately reflect the characteristics of user groups and the patterns of business processing for different electricity businesses, and achieve differentiated representation of high-frequency and low-frequency businesses. By measuring the homogeneity of user groups through the average cosine similarity and combining it with the proportion of processing frequency to reflect the popularity of the business, the user concentration index becomes more comprehensive and objective.

[0019] Furthermore, the recommendation level satisfies:

[0020] In the formula, For the first Recommendation level of each power service for target users Processing the first for target users The number of times a power business is conducted. The concentration of historical dynamic characteristics of the target users. For the first User concentration of individual power businesses It is a linear normalization function.

[0021] The beneficial effects are as follows: By exponentially calculating the normalized value of the number of historical user behaviors and the concentration of dynamic features, and then adding it to the concentration of business users, a non-linear recommendation calculation model is constructed. This model can dynamically adjust according to the stability of user behavior. When the concentration of user behavior is high, the exponential function amplifies the influence of historical user behavior and strengthens the recommendation effect. When the concentration of user behavior is low, the exponential function smooths the data and avoids over-reliance on unstable historical data. By introducing the concentration of business users, the group characteristics of the business itself are comprehensively considered, so that the recommendation results reflect both individual preferences and business characteristics. Linear normalization ensures the comparability of data with different dimensions and improves the accuracy and stability of recommendation calculation.

[0022] Furthermore, the intelligent recommendation for self-service electricity business for enterprise users includes: sorting all electricity properties by recommendation level for the target user from highest to lowest, selecting a preset number of electricity services as the top recommendation for the target user in order, and completing the intelligent recommendation for self-service electricity business for enterprise users.

[0023] The beneficial effects are as follows: By sorting all power services from highest to lowest recommendation level and selecting a preset number of services for top recommendation, the orderly display and accurate push of recommendation results are achieved, providing users with the most relevant service recommendations intuitively and improving the efficiency of users searching for and handling services; by controlling the preset number, both the comprehensiveness of the recommendations are ensured and information overload is avoided, thus improving the user experience; the sorting mechanism ensures that high-priority services are displayed first, making the recommendation results more in line with the actual needs of users and improving the level of intelligence of self-service services.

[0024] The present invention has the following beneficial effects:

[0025] (1) Breaking through the limitations of existing technologies that rely solely on consumer preference information, by clustering the historical dynamic feature vectors generated from multiple user visits, dynamic feature clusters are generated and the concentration of historical dynamic features is calculated. This quantifies the stability (high concentration indicates focused demand) and diversity (low concentration indicates dispersed demand) of user needs, and can capture the dynamic change patterns of user behavior. This avoids recommendation bias caused by the one-sidedness of static preference information, and makes the recommendation results more in line with the user's real needs.

[0026] (2) By calculating the user concentration of the target business, we can accurately distinguish between high-concentration businesses (such as electricity bill payment, which has many users and high frequency) and low-concentration businesses (such as power equipment upgrade applications, which have few users and low frequency). Based on the differences in business characteristics, we can formulate targeted recommendation strategies to avoid the same recommendation pattern. For example, for high-concentration businesses, we can focus on the stable demand of high-frequency users, and for low-concentration businesses, we can focus on the mining of potential demand, which can significantly improve the adaptability of recommendations for different types of power businesses.

[0027] (3) For new users visiting for the first time, although they lack historical dynamic behavior, they can participate in the calculation of business user concentration through static feature vectors at the time of registration (such as using the average value of static feature similarity), which provides a basic recommendation basis for cold start scenarios. Compared with the gap in the existing technology for new user recommendations, the present invention effectively fills the lack of cold start compensation mechanism and expands the coverage of intelligent recommendation.

[0028] (4) The recommendation calculation integrates the user's historical frequency of handling business, the concentration of their own dynamic characteristics (reflecting the stability of demand) and the user concentration of the business (reflecting the universality of the business) to form a non-linear adjustment mechanism. For example, for users with stable demand (high dynamic concentration), the recommendation weight of their high-frequency business can be amplified; for users with diverse demand (low dynamic concentration), the recommendation weight can be smoothed to cover more potential businesses, so that the recommendation results can dynamically adapt to changes in user preferences and improve the flexibility and accuracy of the recommendation. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating the steps of an intelligent recommendation method for self-service electricity business for enterprise users, according to an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions in the embodiments of the present invention will be clearly and completely described below. The described embodiments are only a part of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0032] Please see Figure 1 The diagram illustrates a flowchart of a smart recommendation method for self-service electricity business for enterprise users, provided by an embodiment of the present invention. The method includes the following steps:

[0033] S1: Obtain the static feature vectors of each enterprise user and the historical dynamic feature vectors of each power service access.

[0034] Any enterprise user is designated as the target user. Based on the target user's static characteristics at the time of registration (including enterprise size, electricity type, industry classification, region, and remaining electricity), the static feature vector of the target user is determined (including dwell time, keyword input, and access time period). Based on each historical dynamic characteristic of the target user when accessing each electricity service, the historical dynamic feature vector of the target user when accessing each electricity service is determined.

[0035] Specifically, determining the static feature vector of the target user includes:

[0036] The normalized (e.g., min-max normalization) results of continuous data in the static features of the target user at the time of registration are compared with the encoded categorical data in the static features of the target user at the time of registration (e.g., one-hot encoding; assuming a static feature has 4 categories, for example, electricity usage type is divided into 4 categories: residential, commercial, industrial, and agricultural; the first category is encoded as follows). The second type is encoded as (And so on) The results are combined (e.g., after normalizing the enterprise size, the results are...). The type of electricity consumption is coded as Then the static eigenvector is This yields the static feature vector of the target user.

[0037] Specifically, determining each historical dynamic feature vector of a target user when accessing various power services includes:

[0038] The normalized results of continuous data in each historical dynamic feature of the target user when accessing various power services (including online payment, installation application, bill inquiry, fault reporting, and energy efficiency analysis report application) are combined with the encoded results of categorical data in each historical dynamic feature of the target user when accessing various power services to obtain each historical dynamic feature vector of the target user when accessing various power services.

[0039] S2: Determine the concentration of historical dynamic characteristics of each enterprise user.

[0040] It should be noted that different enterprise users have different business needs for the power service platform, and the realization of different business needs corresponds to different dynamic characteristics. If all the historical dynamic characteristics of an enterprise user are the same, then when the enterprise user visits the power service platform again, the power service corresponding to the historical dynamic characteristics can be placed in the priority position for the enterprise user to choose. However, in reality, for an enterprise user, the dynamic characteristics generated each time they visit the power service platform are not exactly the same, and the power services they handle are not exactly the same. Therefore, it is not possible to simply recommend power services based on the enterprise user's dynamic characteristics. In order to avoid the health check results being out of touch with the user's actual needs, this step obtains the concentration of the enterprise user's historical dynamic characteristics for each power service.

[0041] Cluster all historical dynamic feature vectors of the target user when accessing all power services to obtain multiple dynamic feature clusters of the target user. The concentration of the target user's historical dynamic features is determined based on the maximum and minimum distances between the cluster centers of all dynamic feature clusters of the target user.

[0042] Specifically, the clustering uses the mean-shift clustering algorithm.

[0043] Specifically, the concentration of the historical dynamic features satisfies:

[0044] ;

[0045] In the formula, The concentration of historical dynamic characteristics of the target users. It is the maximum distance among the cluster centers of all dynamic feature clusters of the target user. It is the minimum distance between the cluster centers of all dynamic feature clusters of the target user. for Type curve function.

[0046] in, This represents the difference between the dynamic feature vectors of the target users. A larger value indicates a more dispersed dynamic feature vector distribution among the target users, and thus a lower concentration of their historical dynamic features; a smaller value indicates a more similar dynamic feature vector distribution among the target users, and thus a higher concentration of their historical dynamic features. For ease of subsequent calculations, this is expressed here as... Type curve function pairs Normalization is performed, where, , This is a natural constant. For users accessing the power business platform for the first time, since they do not have historical dynamic feature vectors, their historical dynamic feature vector concentration is set to 0 for ease of subsequent calculations.

[0047] S3: Determine the user concentration of each power business.

[0048] It should be noted that, since different electricity services have different functions and different corporate users have different needs for electricity services, some services may be handled by most corporate users (such as bill payment). For these services, priority should be given to most corporate users. Some services may only be handled by a few types of corporate users. For these electricity services, priority can be given only to these types of corporate users. Therefore, in order to measure which electricity services need to be widely recommended and which electricity services need to be recommended to specific types of corporate users, this step obtains the user concentration of each electricity service based on the historical dynamic characteristics concentration of corporate users and the historical electricity service handling records of corporate users.

[0049] Any electricity service is designated as the target service, and enterprise users who have handled the target service are designated as service users. The user concentration of the target service is determined based on the mean of the historical dynamic feature concentration of all service users, the mean of the cosine similarity between the static feature vectors of all service users, and the proportion of the number of times each service user handles the target service to the total number of times they handle all electricity services.

[0050] Specifically, the user concentration satisfies:

[0051] ;

[0052] In the formula, For the first User concentration of individual power businesses For all those who have processed the first The mean of the historical dynamic characteristics concentration of enterprise users in the power business. For all those who have processed the first The mean cosine similarity among the static feature vectors of enterprise users in the power business. For all those who have processed the first The number of enterprise users in the electricity business For all those who have processed the first The first among the enterprise users of the power business The first enterprise user to process the The number of times a power business is conducted. For all those who have processed the first The first among the enterprise users of the power business The number of times an enterprise user handles all electricity-related services.

[0053] in, The representative handled the first The degree of similarity in static characteristics among enterprise users of a power business; the higher the value, the more likely they have processed the transaction. The more similar the enterprise users of the electricity business are, the better the processing of the first... The more likely a power business's enterprise users are to be concentrated in a certain type of enterprise user, the more likely the first The higher the user concentration of a power business, the smaller the value, indicating that the business has been processed for more than one time. The greater the differences between the enterprise users of the electricity business, the more difficult it is to process the first... The more likely a power business enterprise user is to be of various types, the more likely the first one is to be a third party. The lower the user concentration of a power business. The representative handled the first The degree of similarity in access habits among enterprise users of a power business; the higher the value, the more similar they are to the number of times they have processed [the first / secondary service]. The more similar the access habits of enterprise users in the power business platform, the easier it is to process the first... The more likely a power business's enterprise users are to be concentrated in a certain type of enterprise user, the more likely the first The higher the user concentration of a power business, the smaller the value, indicating that the business has been processed for more than one time. The greater the differences in access habits among enterprise users of the power business platform, the more difficult it will be to process electricity-related matters. The more likely a power business enterprise user is to be a user of various types, the more likely they are to be the first The lower the user concentration of a power business, the better. For example, The weights are all The implementers can adjust the weights according to the actual situation. Representatives who have handled electricity business Enterprise users of electricity services The proportion of cases processed indicates the number of cases processed. The higher the value, the more likely the cases have been processed. The more a corporate user of a particular electricity service tends to choose that particular electricity service when processing business transactions, the more likely they are to choose that service. The higher the user concentration of a power business, the smaller the value, indicating that the business has been processed for more than one time. For enterprise users of electricity services, the less inclined they are to purchase this particular electricity service when processing business transactions, the more likely they are to purchase it. The lower the user concentration of a power business.

[0054] S4: Determine the recommendation level of each power service for each enterprise user.

[0055] It should be noted that the historical dynamic characteristic concentration of enterprise users can reflect users' access habits to the power business platform, while the user concentration of power business reflects the access tendency of various types of enterprise users to different power businesses. Since the business handling of enterprise users is affected by both individual factors (such as electricity bill payment) and overall factors (such as various factors affecting the industry as a whole to handle a certain business), this step obtains the recommendation degree of each power business for each enterprise user based on the user concentration of each power business and the historical dynamic characteristic concentration of each enterprise user.

[0056] The recommendation level of the target service for the target user is determined based on the number of times the target user has processed the target service, the concentration of historical dynamic feature vectors, and the user concentration.

[0057] Specifically, the recommendation level satisfies:

[0058] ;

[0059] In the formula, For the first Recommendation level of each power service for target users Processing the first for target users The number of times a power business is conducted. The concentration of historical dynamic characteristics of the target users. For the first User concentration of individual power businesses It is a linear normalization function.

[0060] in, Representing the target user's view on the first The historical processing frequency of a power service, the higher the value, the greater the target user's interest in the first service. The more biased the handling of individual electricity services, the more likely the first one will be to be affected. The higher the recommendation level of the first electricity service for the target user, the lower the value, indicating that the target user prefers the second service. The less biased the handling of individual electricity services, the better. The lower the recommendation level of a particular power service for the target user, the lower the recommendation level. (This is followed by an unrelated sentence about the concentration of historical dynamic characteristics of the target user.) The larger the value, the more stable the target user's service preferences are. Therefore, it's crucial to recommend frequently used services to the target customer, ensuring that the recommendation rate for frequently used electricity services is higher than that for other electricity services. Thus, through... In the form of an exponential function Adjustments should be made. The larger the value, the more prominent the recommendation of the electricity services that the target user frequently handles is compared to other electricity services. The smaller the user base, the more similar the recommendation rates of various power services for their target users. However, for new enterprise users, due to their new business nature... The value is 0, making it impossible to effectively recommend electricity services. Therefore, through... To supplement the recommendation system, the system recommends the electricity services that enterprise users frequently handle to new target users.

[0061] S5: Based on the recommendation level, intelligent recommendations for power business self-services for enterprise users are realized.

[0062] Specifically, the intelligent recommendation for realizing self-service electricity business for enterprise users includes:

[0063] All power services are ranked from highest to lowest recommendation level for the target user. A preset number of power services are then selected as the top recommendations for the target user, thus completing the intelligent recommendation of power service self-service for enterprise users.

[0064] The implementers can set a preset number based on the specific implementation situation, for example, 3.

[0065] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for intelligent recommendation of self-service electricity business for enterprise users, characterized in that, include: Any enterprise user is designated as the target user. Based on the static characteristics of the target user at the time of registration, the static feature vector of the target user is determined. Based on each historical dynamic characteristic of the target user when accessing each power service, each historical dynamic feature vector of the target user when accessing each power service is determined. Cluster all historical dynamic feature vectors of the target user when accessing all power services to obtain multiple dynamic feature clusters of the target user. Determine the historical dynamic feature concentration of the target user based on the maximum and minimum distances between the cluster centers of all dynamic feature clusters of the target user. Any electricity service is designated as the target service, and enterprise users who have handled the target service are designated as service users. The user concentration of the target service is determined based on the mean of the historical dynamic feature concentration of all service users, the mean of the cosine similarity between the static feature vectors of all service users, and the proportion of the number of times each service user handles the target service out of the total number of times they handle all electricity services. This concentration includes: In the formula, For the first User concentration of individual power businesses For all those who have processed the first The mean of the historical dynamic characteristics concentration of enterprise users in the power business. For all those who have processed the first The mean cosine similarity among the static feature vectors of enterprise users in the power business. For all those who have processed the first The number of enterprise users in the electricity business For all those who have processed the first The first among the enterprise users of the electricity business The first enterprise user to apply for The number of times a power business is conducted. For all those who have processed the first The first among the enterprise users of the power business The number of times an enterprise user handles all electricity-related services; Based on the number of times the target user has processed the target service and the concentration of its historical dynamic feature vector, as well as the user concentration, the recommendation level of the target service for the target user is determined, including: In the formula, For the first Recommendation level of each power service for target users Processing the first for target users The number of times a power business is conducted. The concentration of historical dynamic characteristics of the target users. It is a linear normalization function; Based on the aforementioned recommendation level, intelligent recommendations for self-service electricity business for enterprise users can be achieved.

2. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The static characteristics include enterprise size, electricity consumption type, industry classification, region, and remaining electricity.

3. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The dynamic features include dwell time, keyword input, and access time period.

4. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The determination of the static feature vector of the target user includes: The normalized result of the continuous data in the static features of the target user at the time of registration is combined with the encoded result of the categorical data in the static features of the target user at the time of registration to obtain the static feature vector of the target user.

5. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The determination of each historical dynamic feature vector of a target user when accessing various power services includes: The normalized result of continuous data in each historical dynamic feature of the target user when accessing each power service is combined with the encoded result of categorical data in each historical dynamic feature of the target user when accessing each power service to obtain each historical dynamic feature vector of the target user when accessing each power service.

6. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The clustering uses the mean-shift clustering algorithm.

7. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The concentration of the historical dynamic features satisfies: ； In the formula, The concentration of historical dynamic characteristics of the target users. It is the maximum distance among the cluster centers of all dynamic feature clusters of the target user. It is the minimum distance between the cluster centers of all dynamic feature clusters of the target user. for Type curve function.

8. The intelligent recommendation method for self-service electricity business for enterprise users according to claim 1, characterized in that, The intelligent recommendation system for enabling self-service electricity business for enterprise users includes: All power services are ranked from highest to lowest recommendation level for the target user. A preset number of power services are then selected as the top recommendations for the target user, thus completing the intelligent recommendation of power service self-service for enterprise users.

Images