Developing an AI-empowered Chinese Composition Assessment System for Hong Kong Students

Abstract

With the advancements in natural language processing and the growing adoption of Large Language Models (LLMs) in education, AI-based automated essay evaluation has emerged as a significant topic in research on writing literacy assessment. Despite this, there is limited research on Chinese automated essay scoring systems, leaving the subject largely underexplored. This paper introduces an AI-empowered Chinese composition assessment system specifically designed for primary and secondary schools in Hong Kong. The system features customizable prompts based on marking criteria, employs multiple AI models to enhance grading accuracy, and enables rapid report generation. We will outline the system's design, workflow, and future directions for field testing and eventual implementation.