User label distribution prediction method, device and system

Abstract

One or more embodiments of the invention provide a user label distribution prediction method, device and system. The method comprises the steps that a user label prediction model is obtained in advance based on sample data training, the sample data is generated based on collected image information of a sample user and terminal identification information, and the sample data comprises the corresponding relation between visual attribute distribution and high-dimensional label distribution of the sample user; furthermore, in a user label distribution prediction stage, user visual attribute distribution data is determined based on the image information of the passing user collected on site; the user visual attribute distribution data is taken as input data of the user label prediction model; the user label distribution data of the target area is obtained through prediction, so that only the image information of the user needs to be collected in real time in the label distribution prediction stage, the problem that the privacy of the user is violated due to random detection of the terminal identification information of the user can be avoided, and the user label distribution data can beaccurately predicted.

Description

technical field [0001] One or more of the descriptions relate to the field of Internet technologies, and in particular to a method, device and system for predicting user tag distribution. Background technique [0002] At present, with the change and improvement of living standards, people's outdoor activities have increased, and they can frequently come into contact with various outdoor media when waiting for cars, traffic jams, waiting for elevators, and entering subway stations. Based on this, the offline advertising system has become one of the ways to promote resources to users. By setting up an advertising display screen in a target area with a relatively large flow of people, the promotion resources can be promoted, and the users who pass through the target area will be more interested in advertising. Attention and purchase desire of the promotional resources displayed on the screen. [0003] Among them, in order to improve the accuracy of promotional resource placeme...

AI Technical Summary

Problems solved by technology

At present, it mainly collects the terminal mac address of the user passing through the target area in real time, and then obtains the user's online consumption data corresponding to the terminal mac address, and then determines the crowd label distribution of the target area based on the user's online consumption data. In this method, due to the need to detect the terminal mac addresses of all passing users, there is a problem of violating user privacy

Images