Webpage classification method, terminal equipment and storage medium

Abstract

The invention relates to a webpage classification method, terminal equipment and a storage medium, and the method comprises the steps: S1, collecting various types of webpages, constructing a graph structure according to at least two types of features in each webpage, marking the types of the webpages, and enabling all graph structures with type marks to form a training set; S2, constructing a graph convolutional neural network model, training the graph convolutional neural network model through the training set, and taking the trained model as a webpage classification model; and S3, for the to-be-classified webpage, constructing a graph structure according to the at least two types of features in the step S1, and determining a webpage type corresponding to the graph structure through a webpage classification model. According to the method, additional heterogeneous information such as texts and pictures in the web pages is fully learned to construct the web page classification model, and compared with an existing web page classification method, the limitation of the web page classification method based on a single data structure can be effectively solved, and the problem of web page feature sparseness can be obviously solved.

Description

technical field [0001] The invention relates to the field of webpage classification, in particular to a webpage classification method, terminal equipment and a storage medium. Background technique [0002] With the rapid popularization of Internet technology, Internet applications are also booming. High-quality, personalized content is constantly emerging, and more and more netizens can share rich network resources. But at the same time, some illegal and criminal activities are also hidden in it, and a large amount of false information, advertising information, Internet fraud and other illegal and illegal information are released on the Internet, which seriously endangers the property safety of the majority of netizens. How to discover and identify this kind of bad text information and purify the network space requires an efficient and intelligent web page analysis method. [0003] The content information structure of the web page is diverse, with pictures, texts, videos a...

AI Technical Summary

Problems solved by technology

This method is less flexible and has the problem of low generalization ability; 2) building a classification model based on webpage content, most of the webpage features currently used are too single, such as only using text information or only using image information as the feature representation of webpages, It cannot fully represent the content information of the webpage, ignoring the information carried by other structural data, often ignoring the key information and causing the features to be more sparse, which has obvious limitations

Existing webpage classification methods have the following deficiencies: (1) by comparing webpage content or URLs at present, it is usually necessary to build a large-scale comparison library. High error rate and poor generalization; (2) At present, the classification model is built based on webpage content. Since only single-structure data is considered in the modeling process, but the information structure of webpage content is diverse, some webpages may only have text or pictures, etc.

Therefore, it is easy to see that the classification method based on single-structure data cannot be widely applied to all web page data, and cannot solve the problem of feature sparsity. The scope of application is very limited, and the model effect cannot be guaranteed.

Images