The usage of artificial Intelligence-empowered text analysis model with convolutional neural network in english...

Area of Science:

• Educational Technology
• Artificial Intelligence in Education
• Natural Language Processing

Background:

• Personalized reading instruction is crucial for student success.
• Current AI applications in education offer potential but require refinement for specific subjects like high school English.
• The need for effective tools to support both students and educators in reading comprehension is significant.

Purpose of the Study:

• To systematically apply the Text Convolutional Neural Network (Text CNN) model in real high school English classrooms for personalized reading instruction.
• To develop and evaluate a teaching-assistive framework integrating dynamic content recommendation and real-time feedback.
• To assess the impact of the Text CNN model on student reading comprehension and teacher workload.

Main Methods:

• Development of a Text CNN-based framework with multi-scale feature extraction and an attention mechanism.
• Automatic classification of reading materials, extraction of key terms, and personalized resource delivery.
• A comparative experiment involving 60 high school students (experimental vs. control group).

Main Results:

• The experimental group using the Text CNN tool showed a significant improvement in average reading comprehension scores (66.2 to 79.6).
• The Text CNN model achieved high classification accuracies: 93.5% for science texts and 94.2% for education texts.
• Teacher grading workload was reduced by an average of 70% through the AI tool.

Conclusions:

• The Text CNN model provides empirical evidence for its effective application in English reading instruction.
• The developed framework offers a practical foundation for intelligent educational tools, enhancing personalized learning.
• Further research is needed to explore the model's performance on diverse and unstructured text types.