Development of a convolution algorithm for the content of a Web resource for processing by a neural network

Abstract

The article aim to develop a convolution algorithm for the content of Web resources for processing by a neural network. The content of a web resource is represented in a vector form based on a co-occurrence matrix. A convolutional neural network, where the most significant features stand out from them, processes the content vectors of Web resources. The result is a collapsed content of Web resources in the form of a vector of significant features, which is the result