A Study on the Readability of Chinese Text from a Rhetorical Perspective

Text readability calculation is an important method to evaluate the difficulty of Chinese reading text. This paper based on rhetoric corpus, counted the rhetoric features related index, adopt quantitative means to analyze the correlation of rhetoric features and text readability and other level language features fusion, and then use the multiple linear regression analysis method, with each level text features related index as an independent variable, the text difficulty level as dependent variable, to construct text readability calculation formula model, and calculated the model effect fitting degree. The results of this paper show that there is a strong correlation between the proportion of common rhetorical features, rhetorical questions, references and the readability of the text, which are more than 90%. The model constructed in this paper has a high fitting degree (R2 = 0.972). In addition, the text test performs well in low-level text, and the prediction accuracy of high-level text is not enough.