A method, device and equipment for evaluating web pages

Abstract

The purpose of the invention is to provide a method, a device and equipment for confirming a webpage evaluating model based on a transfer learning method. The method comprises the steps as follows: obtaining a set of target sample sets and at least one set of reference sample sets corresponding to the target sample sets in a webpage evaluating model to be established; obtaining weight adjustment information of part or all of the reference sample sets relative to the target sample sets in the at least one set of reference sample sets according to the target sample sets and the at least one set of reference sample sets; and performing a transfer operation according to the predetermined training mode and the weight adjustment information corresponding to the target sample sets and the reference sample sets to obtain the webpage evaluating model of the target sample sets. The method has the advantages as follows: the convergence operation can be quickly performed to establish the stable evaluation model, so that the method can be satisfactorily suitable for processing mass data and can obtain a better effect for evaluating current mass web pages.

Description

technical field [0001] The present invention relates to the field of computer technology, in particular to a method, device and equipment for evaluating webpages. Background technique [0002] As the Internet is increasingly integrated into people's lives, Internet advertising has become an important means for companies to carry out product / service marketing. Correspondingly, users urgently need to know how effective advertising is, so as to adjust advertising strategies accordingly . However, it is not only inefficient to use all the data obtained to build a model each time, but also the deviation and variation of the model are relatively large. Therefore, it is an urgent need for many people to establish an evaluation model for evaluating the advertisement information to be placed based on known data, that is, to establish an evaluation model by adopting a migration method. [0003] However, due to the rapid expansion of the scale of the Internet, the amount of various t...

AI Technical Summary

Problems solved by technology

However, the existing migration algorithm needs to be implemented through iterative operations on each data. In the case of massive data, it is difficult to achieve a stable convergence state even through a long period of learning and training operations.

Images