Using Conventional Neural Networks for the Problem of Text Classification

Abstract

Convolutional neural networks are a powerful machine learning tool that aims to efficiently recognize and classify images. The success of using convolutional neural networks for images has spawned many attempts to use this tool in other tasks. In this paper, the main methods of using convolutional neural networks for the text classification problem are investigated. Experiments on large text data have been carried out, showing that convolutional neural networks for the text classification problem can achieve a quality similar to or better than traditional methods.