Author(s)

Tianhuan Ma

Corresponding Author

Tianhuan Ma

Abstract

This paper analyzes the basic research and application of Chinese essay automatic scoring in recent 20 years (2000-2020) in China, finding that on the whole, the Chinese essay automatic scoring system basically reaches the same level as human scorers, and the existing scoring system has been put into use in large-scale essay examination scoring in some areas. As is reflected by basic research and the working principle of machine scoring, the automatic evaluation on the language quality of essays has achieved good results, but the quality of machine-based evaluation on the semantic content and textual structure is far from satisfactory, and the underlying motivations are not that justifiable. One reason is that the basic studies on essay scoring are small in number, and another reason is that the machine-driven analysis of Chinese textual semantics remains defective. Therefore, it is crucial to realize the automation evaluation on the semantic content and textual structure of Chinese essays.

Keywords

Chinese, Automated scoring, Essay