Target user acquisition method, device and server

Abstract

The invention provides a target user obtaining method, a device and a server. The method includes obtaining a user label directed graph, wherein the user label directed graph takes a label held by a user as a vertex, so that a directed line segment of a first label pointing to a second label indicates a probability that a user holding the first label simultaneously holds the second label; the userlabel directed graph obtains a user label directed graph, and the user label directed graph obtains a user label directed graph. Acquiring an original identification directed graph corresponding to the user tag directed graph; The preset clustering algorithm updates the identification of each vertex in the original digraph iteratively, and records the identification update sequence of each vertexin the clustering process. A target us is acquired accord to that identification update sequence. The invention can provide a target crowd obtaining method capable of accurately locating a target crowd, simultaneously taking into account complexity and not limited to the number of original users of a tag, and the target user obtaining method has good accuracy.

Description

technical field [0001] The present invention relates to the field of computers, and in particular, to a method, device and server for acquiring a target user. Background technique [0002] With the development of Internet technology, people are increasingly relying on the Internet to obtain information on life, entertainment and work. There are a large number of target group acquisition schemes, and the tag group acquisition scheme aims to acquire the delivery requirements of information publishers, and acquire target groups that accept the information according to user tags. However, existing tag crowd acquisition schemes are largely limited by the number of original users holding tags, and cannot acquire larger-scale target groups. [0003] In order to solve this technical problem, a crowd diffusion scheme is proposed in the prior art. The current mainstream crowd diffusion scheme is to transform the crowd diffusion problem into a binary classification problem, but this ...

AI Technical Summary

Problems solved by technology

[0004] (1) It is necessary to build a classifier for each label, and the cost of training and maintenance is very high;

[0005] (2) If there are long-tail unpopular tags, the number of users is too small, which will affect the accuracy of diffusion

Images