Investigating genre distinctions through discourse distance and discourse network

Abstract

The notion of genre has been widely explored using quantitative methods from both lexical and syntactical perspectives. However, discourse structure has rarely been used to examine genre. Mostly concerned with the interrelation of discourse units, discourse structure can play a crucial role in genre analysis. Nevertheless, few quantitative studies have explored genre distinctions from a discourse structure perspective. Here, we use two English discourse corpora (RST-DT and GUM) to investigate discourse structure from a novel viewpoint. The RST-DT is divided into four small subcorpora distinguished according to genre, and another corpus (GUM) containing seven genres are used for cross-verification. An RST (rhetorical structure theory) tree is converted into dependency representations by taking information from RST annotations to calculate the discourse distance through a process similar to that used to calculate syntactic dependency distance. Moreover, the data on dependency representations deriving from the two corpora are readily convertible into network data. Afterwards, we examine different genres in the two corpora by combining discourse distance and discourse network. The two methods are mutually complementary in comprehensively revealing the distinctiveness of various genres. Accordingly, we propose an effective quantitative method for assessing genre differences using discourse distance and discourse network. This quantitative study can help us better understand the nature of genre.

Tell me how you write and I'll tell you what you read: a study on the writing style of book reviews

“…Despite the long-standing and widespread attention, the AGC task is extremely challenging due to multiple factors. Previous works differ with respect to 1) the notion of genre, which represents a quite heterogeneous object of study (Biber and Conrad, 2009), 2) the typologies of text properties, either token- (Crossley and Louwerse, 2007;HaCohen-Kerner et al, 2020;Mehler et al, 2010;Santini, 2004) or sentence-based characteristics (Cimino et al, 2017;Fang and Cao, 2010;Stamatatos et al, 2001;Wan et al, 2019) or also accounting for the discourse structure of texts (Sun et al, 2021), 3) the machine learning approach adopted (Worsham and Kalita, 2018), and 4) the source exploited to classify the genre, either the book content (Rahul et al, 2021;Shamir, 2020;Worsham and Kalita, 2018), title or summary (Ozsarfati et al, 2019), descriptions on websites (Sobkowicz et al, 2018), or cover design (Buczkowski et al, 2018).…”