A Multi-Dimensional Framework and Practical Approach for Leveling Chinese Texts for Primary School Students

Abstract

Text difficulty is not only influenced by linguistic complexity but is also closely related to factors such as readers’ cognitive load and cultural context. Current readability formulas have limitations, including overreliance on surface-level text features and neglect of reader differences. Therefore, this paper proposes that when leveling Chinese texts for primary school students, four dimensions should be considered: language, content, cultural context, and genre structure. By combining natural language processing technology to extract features from multiple dimensions such as vocabulary, sentence structure, content, and structure, a text difficulty norm applicable to different grade levels can be established. Through a combination of expert evaluation and data comparison, the difficulty level of the text can be accurately determined.