Analysis of Intelligent Recommendation System for Literature Education Using Deep Learning

Abstract

Teaching literature is essential for fostering critical thinking, creativity, empathy, and cultural understanding among pupils. However, traditional methods face challenges, including incomplete instruction, limited access to diverse resources, and insufficient engagement, particularly in poetry education. To address these issues, this study proposes an Advanced Sand Cat Swarm-driven Consecutive Convolutional Neural Network (ASCS-CCNN) approach that enhances poetry fluency and literature strategies. The model integrates a recommendation system using collaborative filtering to optimize the sequencing of educational resources. Data preprocessing involves tokenization and feature extraction using Term Frequency-Inverse Document Frequency (TF-IDF). The proposed ASCS-CCNN method evaluates students’ performance in poetry instruction based on metrics like similarity, tone correctness, and rhyme accuracy. The findings reveal that incorporating well-chosen texts aligned with curriculum objectives significantly improves student engagement and appreciation of poetry. The ASCS-CCNN model demonstrates superior performance, achieving enhancements in consistency (4.88%), fluency (5.02%), meaningfulness (4.94%), and poutiness (4.82%). This innovative approach offers a promising solution to improve literature education for primary and secondary students.