Transformer-based Context-aware Sarcasm Detection in Conversation Threads from Social Media
Computation and Language
2020-05-26 v1
Abstract
We present a transformer-based sarcasm detection model that accounts for the context from the entire conversation thread for more robust predictions. Our model uses deep transformer layers to perform multi-head attentions among the target utterance and the relevant context in the thread. The context-aware models are evaluated on two datasets from social media, Twitter and Reddit, and show 3.1% and 7.0% improvements over their baselines. Our best models give the F1-scores of 79.0% and 75.0% for the Twitter and Reddit datasets respectively, becoming one of the highest performing systems among 36 participants in this shared task.
Cite
@article{arxiv.2005.11424,
title = {Transformer-based Context-aware Sarcasm Detection in Conversation Threads from Social Media},
author = {Xiangjue Dong and Changmao Li and Jinho D. Choi},
journal= {arXiv preprint arXiv:2005.11424},
year = {2020}
}
Comments
To be published in ACL FigLang2020