Application and evaluation system of deep learning in automatic summary generation of English reading materials

With the rapid development of the field of artificial intelligence, deep learning has solidified its position as a major technology in the field of computational linguistics. The direction of this academic effort is to investigate the utility of deep learning in automating the creation of abstracts of literary works in English, while developing a critical mechanism for assessing their proficiency. Through careful study and comparative analysis of different deep learning paradigms, including recurrent neural networks (RNNS), Long short-term memory networks (LSTMs), and the cutting-edge Transformer model, we conceived of a streamlined automated profile generator. With the constraints and validation of a large number of English texts, the results show that the Transformer driven model achieves higher accuracy in terms of accuracy, narrative fluidity, and information density contained in the summary.