Syntax and Coherence - The Effect on Automatic Argument Quality Assessment.

In this paper, we focus on the task of automatic argument quality assessment. Prior empirical methods largely ignore the syntax structure in one argument or depend on handcrafted features that have shallow representation ability. In contrast, we proposed a method that directly models syntax and topic coherence. Our method can acquire both topic coherence and syntactic information from an argument that explicitly utilizes various types of relationships among words, thus can help with argument quality assessment. Experimental results suggest that our method significantly outperforms the previous state-of-the-art method and strongly indicates syntax and coherence correlate with argument quality.