Style Matters! Investigating Linguistic Style in Online Communities

TL;DR

This study explores the linguistic styles of online communities, demonstrating that each community has a distinct style that can accurately predict group membership and is resilient to less training data.

Contribution

It introduces a broad, 262-feature definition of linguistic style and shows that community styles are distinguishable and predictive of group membership.

Findings

Communities have distinct linguistic styles.

Style predicts community membership with high accuracy.

Style remains effective even with reduced training data.

Abstract

Content has historically been the primary lens used to study language in online communities. This paper instead focuses on the linguistic style of communities. While we know that individuals have distinguishable styles, here we ask whether communities have distinguishable styles. Additionally, while prior work has relied on a narrow definition of style, we employ a broad definition involving 262 features to analyze the linguistic style of 9 online communities from 3 social media platforms discussing politics, television and travel. We find that communities indeed have distinct styles. Also, style is an excellent predictor of group membership (F-score 0.952 and Accuracy 96.09%). While on average it is statistically equivalent to predictions using content alone, it is more resilient to reductions in training data.