Automated Scoring and Personalized Feedback Generation for English Writing Through Semantic and Syntactic Complexity Analysis

With the continuous advancement of artificial intelligence technology, automatic writing assessment (AWE) systems have been widely used in English writing teaching. However, existing systems mostly focus on surface features and neglect in-depth analysis of syntactic complexity and semantic levels. This study innovatively proposes a method that combines grammatical complexity and semantic analysis, aiming to improve the accuracy of English writing assessment and achieve personalized feedback. This method evaluates language fluency, logical rigor, and clarity of expression by examining syntactic structure and semantic coherence, and generates more accurate ratings and customized feedback. This study combines deep learning models, syntax tree parsing, and semantic embedding techniques to develop a system that can automatically evaluate English writing and provide targeted recommendations. Experimental results have shown that the system surpasses existing benchmarks in terms of scoring accuracy and feedback practicality, effectively guiding students to improve and promoting the enhancement of writing skills.