Cognitive Psychology Based Text Analysis Using Feature Extraction and Classification by Deep Learning Architectures for English Language Analysis

Abstract

Deep learning can provide English learners an edge and benefit teachers and students. Writing, translating, hearing, and speaking assist English language instruction change. This study proposes a unique deep learning architecture-based cognitive psychology-based educational data training method for English language learning. Student language learning and English teaching data has been collected here. Data is obtained from English-taught language students. Classification and feature extraction are done on this data. This data was extracted using probabilistic Laplacian Score-based Restricted Boltzmann machine. Rough Set Theory-based capsule Net Convolutional neural networks classified these collected features. Educational datasets from students were analysed experimentally for English teaching language learning. Parametric analysis showed Word Perplexity of 90.5%, Flesch-Kincaid (F-K) Grade Level for Readability of 55.6%, Cosine Similarity for Semantic Coherence of 85.4%, gradient change of NN of 46.5%, validation accuracy of 98%, and training accuracy of 96.8 by proposed Pr-Lap Sc RBM-RsT CapsNet CNN.