Automated Assessment of Text Complexity through the Fusion of AutoML and Psycholinguistic Models

The complexity of written texts poses significant challenges for comprehension, impacting education, literacy, and communication across various fields. As the demand for advanced text assessment tools grows, this study aims to integrate Automated Machine Learning (AutoML) with psycholinguistic models to enhance the automated assessment of text complexity, ultimately improving educational practices and content development. A mixed-methods approach combined the quantitative analysis of text complexity metrics with qualitative insights from psycholinguistic models. The AutoML framework automated model selection and hyperparameter tuning, while psycholinguistic features were extracted to inform the model. This research addresses a critical gap in existing automated text assessment methods, which often lack a nuanced understanding of language complexity and rely on simplistic heuristics that fail to capture the intricacies of language. Integrating AutoML and psycholinguistic models offers a more accurate, efficient, and contextually relevant assessment of text complexity, which is crucial for educational tools and content creation. The fusion model achieved an impressive 92% accuracy, outperforming traditional models (77%) and large language models (82%), while demonstrating a rapid response time of 0.5 s, making it suitable for real-time applications. These findings highlight the significant potential of combining AutoML with psycholinguistic insights to enhance automated text complexity assessment. This innovative approach paves the way for improved educational outcomes and more effective communication strategies, offering a promising solution to the challenges of text complexity evaluation in various domains.