A Construction Method of Consumption Portrait of Telecom Users Oriented to Precision Marketing

Abstract

The invention relates to a method for constructing a consumption portrait of a telecommunication user oriented to precision marketing, comprising the following steps: performing data preprocessing on user bill data to obtain user consumption behavior data; using the user bill data and new package charging rules as offline The billing engine inputs, calculates the consumption amount of the user under the new package, uses the consumption amount under the old and new packages to calibrate the class label for the user; trains the random forest model of the new package; screens important features according to the random forest model; for each important feature, calculates the new Split points of non-potential users and potential users of the package; use important features and corresponding split points to construct user portraits of potential users. The invention uses an offline billing engine to calculate the consumption amount of the user under the new package, labels the user, and uses a random forest model to screen important features and construct a user portrait, which is beneficial to the precise marketing of the new package.

Description

technical field [0001] The present invention relates to the fields of data analysis and data mining, and more specifically, to a method for constructing consumption portraits of telecommunications users for precision marketing. Background technique [0002] With the development of the Internet, there are more and more mobile phone applications, and social software emerges in an endless stream, and has become the main way for people to connect. This has had a huge impact on operators' traditional services such as calls and short messages, and operators must open up new development paths. But attracting users is the premise of all new development paths, and operators, as providers of basic communication services, should start with basic communication services to attract and retain users, that is, user package choices. [0003] At present, there are many types of packages offered by operators, and Shanghai Telecom has hundreds of packages. It is difficult for users to choose a...

AI Technical Summary

Problems solved by technology

However, the current user portrait construction methods for telecom users are generally based on user basic attribute data, online behavior data, etc., and the constructed user portrait reflects the basic characteristics of the user; since there is no user consumption bill data, the constructed user portrait Portraits cannot reflect user package consumption

Therefore, the existing methods for constructing telecom user portraits cannot understand the user's package consumption data, and cannot be used for precise marketing of package push

Images