Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A method for mobile advertising platform to find similar users

A similar user and mobile advertising technology, applied in the field of mobile Internet, can solve the problems of different similar users and uncertain clustering results

Active Publication Date: 2018-08-10
GUANGZHOU YOUMI INFORMATION TECH
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the case that the customer has strict requirements on the expansion time of similar users, neither of the above two methods can achieve expansion in a relatively short period of time
For most clustering methods, the clustering results are often uncertain, resulting in different similar users expanded each time by the same batch of seed users.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A method for mobile advertising platform to find similar users
  • A method for mobile advertising platform to find similar users
  • A method for mobile advertising platform to find similar users

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0038] A method for a mobile advertising platform to find similar users, comprising the following steps:

[0039] (1) The developer (advertiser) of the target App submits a list of existing seed user device numbers of the target App;

[0040] (2) Obtain a list of non-similar user device numbers of the target App:

[0041] a. The developer of the target App directly submits a list of non-similar user device numbers;

[0042] b. Randomly extract device numbers equivalent to the similar user list from the advertising platform's own device list, and use it as a list of non-similar user device numbers;

[0043] (3) Utilize the system-level API to obtain the App installation package list of the mobile user;

[0044] (4) installation package filtering: calculate the device coverage rate of each App of mobile users, and remove very high and very low Apps with very high and very low coverage device ratios from the App installation package list; in step (4), the threshold M=50% , thr...

Embodiment 2

[0061] Such as figure 1 , a method for a mobile advertising platform to find similar users, comprising the following steps:

[0062] First, an L2 regularized logistic regression model is trained based on the installation list and tags filtered by the training users. For a new user (see the rounded rectangle for features) installation list, use the trained logistic regression model to get a prediction value between [0,1], indicating the probability of being a similar user. Then calculate the number of paid applications in the user's installation list, the proportion of basic applications, and the average paid price characteristics, combine these characteristics with the results of the logistic regression model in the previous step, and then train a GBDT model to finally predict whether the user is a similar user (1 for similar users, 0 for non-similar users).

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

A method for finding similar users on a mobile advertising platform disclosed by the present invention comprises the following steps: firstly, an L2 regularized logistic regression model is trained according to the filtered installation list and tags of trained users. For a new user installation list, use the trained logistic regression model to get a prediction value between [0,1], indicating the probability that it is a similar user. Then calculate the number of paid applications in the user's installation list, the proportion of basic applications, and the average paid price characteristics, combine these characteristics with the results of the logistic regression model in the previous step, and then train a GBDT model to finally predict whether the user is a similar user . The method of the present invention accurately expands similar users according to the seed users provided by the client under a small amount of calculation.

Description

technical field [0001] The invention relates to the field of mobile Internet, in particular to a method for searching similar users on a mobile advertising platform. Background technique [0002] In the prior art, some solutions are also proposed for searching similar users of the mobile advertising platform. [0003] For example, in Audience segment expansion using distributed in-database k-means clustering (ADKDD2013), it is mainly realized through the following technical solutions: by extracting user-related keywords or topic models as the characteristics of each user, and then using the machine Clustering methods in learning (such as k-means) cluster user groups. Other users in the cluster where the seed user is located are scalable users. [0004] Another example, a similar user search system and method for a video website (application number: 201510142618.6 application date: 2015-03-27), in this technical solution, it is mainly realized through the following technica...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): G06Q30/02G06K9/62
CPCG06Q30/0255G06Q30/0277G06F18/24323
Inventor 李百川陈第李展铿蔡锐涛甄勇
Owner GUANGZHOU YOUMI INFORMATION TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com