Evaluation of Reading Difficulty Based on Factor Analysis and Linear Regression

This paper conducts research on how to measure the difficulty of English texts and extend it other languages, such as Chinese. Firstly, text features that can represent reading are extracted represented using parameters. Secondly, indicators subjected KMO Bartlett's tests, results show these suitable for factor analysis. Then, a comprehensive evaluation model language based analysis is constructed, coefficient composite score F 16 calculated. Additionally, linear regression equation constructed separately correlation between Type-Token Ratio (TTR) F. Finally, conclusions obtained from compared with officially published coefficients in postgraduate entrance examination, validating rationality results. Furthermore, above multi-indicator extended languages Chinese, confirming universality model, effectively solving problem "measuring difficulty.".