Automated Scoring System Using Dependency-Based Weighted Semantic Similarity Model

Abstract

Traditionally, automated scoring system uses semantic similarity between words and the weight of words to calculate semantic similarity between student's answer and standard answer. It doesn't consider the word-order or syntactic information, which can improve the knowledge representation and therefore lead to better performance. This article presents a novel approach called dependency-based weighted semantic similarity model which takes syntactic relations into account and incorporates word-based information in addition to dependency parsing. The experiment shows that compared with traditional word-based weighted semantic similarity model, the dependency-based weighted semantic similarity model improves the precision obviously. It also provides better discrimination of syntactic-semantic knowledge representation than the traditional one.