Resident user portraying method based on multi-dimensional fine-grained behavior data

Abstract

The invention discloses a resident user portraying method based on multi-dimensional fine-grained behavior data. The method comprises four parts: 1, acquiring multi-dimensional fine-grained behavior data, including fine-grained electricity consumption behavior measurement data acquired based on a non-home-entry terminal, user electricity charge data acquired based on a marketing system, and user network behavior statistical data acquired based on an online business hall and 95598; 2, constructing a feature tag model, establishing a user multi-source feature tag system from three dimensions ofuser behaviors, power utilization characteristics and consumption habits, and giving a calculation method or an estimation method of each feature; 3, calculating comprehensive indexes of various characteristic indexes according to seasons and time periods, proposing an improved k-means clustering algorithm, and dividing different power customers into clusters with different attributes by using theimproved k-means clustering algorithm; and 4, visually presenting a user portrait result to serve as a basis for regulating and controlling accurate positioning of the target user.

Description

technical field [0001] The invention relates to a resident user portrait method based on multi-dimensional fine-grained behavior data, and belongs to the technical field of electricity utilization. Background technique [0002] In the face of high peak-to-valley differences, traditional power dispatching and control methods are difficult to deal with. Calling demand-side resources is an important solution. The key basis for the implementation of this measure is the analysis of users' refined characteristics, that is, user portrait research. In layman's terms , user portrait is a tagged user model abstracted based on information such as user social attributes, living habits, and consumption behavior. The user portrait realizes the explicitness of the user's hidden characteristics, assists in insight into user needs, discovers target users, and taps user control potential, thereby improving marketing strategies and grid auxiliary services. [0003] In the early stage of the r...

AI Technical Summary

Problems solved by technology

Overall, the existing research has the following deficiencies: 1) Existing applications focus more on conventional marketing services, such as electricity expenses, user perception, etc., and rarely meet grid-based requirements; 2) Lack of data sources, Most of the data on behavioral habits are obtained through questionnaire surveys and project experiments, which are difficult to implement and the sample size is small, making it difficult to cover all typical users; Taking into account the user's own electricity consumption differences with the calendar and time

Images