Content pushing method, device and computer equipment

Abstract

The present invention provides a content pushing method, device and computer equipment. A method for pushing content, comprising the following steps: obtaining a tag set of a user; obtaining the click rate of each tag in the tag set to an application program; obtaining the Pearson correlation coefficient between each tag in the tag set and other tags; according to the The click-through rate of each tag in the tag set to the application program and the Pearson correlation coefficient between the tag and other tags to obtain the weighted average of all tags in the tag set, and use the weighted average as the user's response to the application. The estimated click-through rate of the program; push content to the user according to the estimated click-through rate of the application program. The content push method combines the Pearson correlation coefficient between user tags and the click-through rate of the user tag to the application to estimate the click-through rate of the application, which improves the accuracy of the application click-through rate estimation, and then can be based on the application. Estimated click-through rate to accurately push content to users.

Description

technical field [0001] The present invention relates to the technical field of the Internet, in particular, the present invention relates to a content pushing method, device and computer equipment. Background technique [0002] In a mobile application (APP) store, a traditional method for estimating the click-through rate of a mobile application is: combining artificial feature engineering and a logistic regression algorithm to estimate the click-through rate of a mobile application. However, this approach has two insurmountable drawbacks. First of all, the features in artificial feature engineering are artificially screened. This approach is too subjective, and it is easy to screen out a large number of useless features to affect the accuracy of the subsequent model. Secondly, the logistic regression algorithm is a linear model, and it is difficult to express a model with nonlinear characteristics. Therefore, the traditional method for estimating the click-through rate of...

AI Technical Summary

Problems solved by technology

However, this method has two insurmountable drawbacks

First of all, the features in artificial feature engineering are artificially screened. This approach is too subjective, and it is easy to screen out a large number of useless features to affect the accuracy of the subsequent model.

Secondly, the logistic regression algorithm is a linear model, it is difficult to express the model of nonlinear characteristics

Therefore, the traditional method for estimating the click-through rate of mobile phone applications cannot accurately estimate the click-through rate of mobile phone applications, and thus cannot accurately push content to users according to the traditional method for estimating the click-through rate of mobile phone applications.

For example, according to the traditional mobile application click rate estimation method, it is not possible to accurately push the application to the user, etc.

Images