Investigation of Deep Learning Algorithms to Reduce the Text Complexity

Abstract

Converting complex texts to simple texts and reducing the complexity of scientific articles are among the issues that have received more attention with the development of deep learning techniques. The simplification of texts greatly helps to expand the dissemination of knowledge. Deep learning semantic similarity algorithms are used to recognize words vector with high accuracy. In this research, by introducing machine learning algorithms such as word2vec, glove, fast text, Bert and Elmo, we will introduce the best algorithm with high efficiency in recognizing and solving the complexity of texts.