Contrastive Learning of Sociopragmatic Meaning in Social Media

TL;DR

This paper introduces a contrastive learning framework specifically designed to capture sociopragmatic meanings in social media, improving performance on diverse sociopragmatic tasks across various datasets and settings.

Contribution

It presents a novel, task-agnostic contrastive learning approach that effectively models sociopragmatic meaning, outperforming existing models in both in-domain and out-of-domain scenarios.

Findings

Achieves 11.66 average F1 point improvement on 16 datasets with minimal training data.

Outperforms other contrastive learning frameworks in diverse sociopragmatic tasks.

Effective in both general and few-shot learning settings.

Abstract

Recent progress in representation and contrastive learning in NLP has not widely considered the class of \textit{sociopragmatic meaning} (i.e., meaning in interaction within different language communities). To bridge this gap, we propose a novel framework for learning task-agnostic representations transferable to a wide range of sociopragmatic tasks (e.g., emotion, hate speech, humor, sarcasm). Our framework outperforms other contrastive learning frameworks for both in-domain and out-of-domain data, across both the general and few-shot settings. For example, compared to two popular pre-trained language models, our method obtains an improvement of $11.66$ average $F_1$ on $16$ datasets when fine-tuned on only $20$ training samples per dataset.Our code is available at: https://github.com/UBC-NLP/infodcl