Pith. sign in

REVIEW

Contrastive Learning of Sociopragmatic Meaning in Social Media

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2203.07648 v5 pith:JE4ND7WM submitted 2022-03-15 cs.CL cs.AI

Contrastive Learning of Sociopragmatic Meaning in Social Media

classification cs.CL cs.AI
keywords learningcontrastivemeaningsociopragmaticframeworklanguageacrossavailable
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original 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

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.