标签设计方面，从电力企业特定业务需求出发，通过规则归纳、数据挖掘等方法

In terms of label design, starting from the specific business needs of electric power companies, the label library is defined to identify customer labels through rule induction, data mining and other methods, and natural language refinement is used to form customer labels; label application is based on the closed-loop circulation of business, data and labels. , Iterative optimization; In terms of label evaluation, evaluate the labels in the application, including the number of times the label is used, the user, the time of use, and the push status, to provide a reference for the update and optimization of the label; to implement the automatic label evaluation in a fixed cycle operation mode; Similar tags, evaluate the importance and real-timeness of tags by customizing tag weights

Label design, from the specific business needs of power enterprises, through the rules of generalization, data mining and other methods to define the label library to identify customer labels, natural language refining, the formation of customer labels; The importance and real-time nature of labels are evaluated in a way that customizes label weights for similar labels.

In terms of label design, starting from the specific business needs of electric power enterprises, the tag library is defined through rule induction, data mining and other methods to identify customer tags, and natural language extraction is used to form customer labels; in the aspect of label application, closed loop and iterative optimization are carried out according to business, data and label; in terms of label evaluation, the label in application is evaluated, including label application The usage times, users, usage time and push situation provide reference for label updating and optimization; the tag automatic evaluation is realized by fixed cycle operation mode; the importance and real-time performance of tags are evaluated by self-defined tag weight for similar tags<br>