English

A Large Self-Annotated Corpus for Sarcasm

Computation and Language 2018-03-26 v4 Artificial Intelligence Machine Learning

Abstract

We introduce the Self-Annotated Reddit Corpus (SARC), a large corpus for sarcasm research and for training and evaluating systems for sarcasm detection. The corpus has 1.3 million sarcastic statements -- 10 times more than any previous dataset -- and many times more instances of non-sarcastic statements, allowing for learning in both balanced and unbalanced label regimes. Each statement is furthermore self-annotated -- sarcasm is labeled by the author, not an independent annotator -- and provided with user, topic, and conversation context. We evaluate the corpus for accuracy, construct benchmarks for sarcasm detection, and evaluate baseline methods.

Keywords

Cite

@article{arxiv.1704.05579,
  title  = {A Large Self-Annotated Corpus for Sarcasm},
  author = {Mikhail Khodak and Nikunj Saunshi and Kiran Vodrahalli},
  journal= {arXiv preprint arXiv:1704.05579},
  year   = {2018}
}

Comments

6 pages, 4 Figures. To Appear in LREC 2018

R2 v1 2026-06-22T19:20:54.620Z