Abstract: In the context of smart education and AI-enhanced teaching, this study developed an intelligent scoring system for physical chemistry experiment reports and implemented it in teaching practice. The system enables automated evaluation and intelligent analysis of student reports while providing learning analytics feedback. Through two semesters of practical application, the system has demonstrated significant improvements in grading efficiency, reduced feedback turnaround time, and maintained high levels of objectivity and accuracy in assessment. This paper details the system's design principles, architecture, and practical implementation, along with an evaluation of its performance and future prospects. The findings offer valuable insights for applying artificial intelligence in higher education, particularly in STEM experimental instruction.

Key words: Physical chemistry experiment, Artificial intelligence technology, Intelligent scoring system for experiment reports, Smart education