Integrating AI for Intelligent Scoring in Junior High School Physics and Chemistry Experimental Operations

This paper explores the application of artificial intelligence (AI) in the intelligent scoring of junior high school physics and chemistry experimental operations. As part of Shanghai's education reform, we investigate the integration of AI technologies, proposing a system built with a multi-modal foundation model and specialized plugin models. Our analysis of experimental results from multiple years of implementation in Shanghai highlights improvements in accuracy and efficiency over traditional processes. The findings indicate that AI-driven scoring can provide reliable and fair assessments, offering valuable insights for future educational examination reforms. This study aims to contribute to the development of high-quality education examinations and support the strategic goals of building a strong education nation. These AI models are integrated into the national AI open innovation platform for educational assessments, offering middleware, SDKs, APIs, and AI cloud services to support various educational scenarios, ultimately providing superior products and services to educational enterprises.