Home // eKNOW 2023, The Fifteenth International Conference on Information, Process, and Knowledge Management // View article

A Corpus Study with German Data Sets into the Similarity of Irony and Satire

Authors:
Marisa Schmidt
Karin Harbusch

Keywords: Satire; Irony; Fake News; Deception Detection

Abstract:
In deception detection, i.e., the falsification of news, satire detection is an import research area. This work strives for high accuracy in satire detection. We want to answer the question whether irony detection can serve the purpose of satire detection as well, or even better than specialized satire classification. The hypothesis underlying this claim follows the definition that satire is a genre that uses irony. Thus, we argue that irony should be indicative in a satire dataset. We contrast the results of runs with irony and satire annotated corpora with Elmo4Irony, an existing classifier for irony, and Adversarial Sat- ire, an existing system for satire detection. In our evaluation, we use three different German data sets labeled with irony and sat- ire, respectively. Our study corroborates the claim. Irony can indeed be found in a satirical dataset—even with higher accu- racy. In order to supplement the finding, both systems are eval- uated with typical examples from satire papers for deeper ex- ploration. Unexpectedly, for the examples from the scholarly lit- erature, both systems can hardly distinguish between irony/sat- ire and neutral formulations.

Pages: 40 to 43

Copyright: Copyright (c) IARIA, 2023

Publication date: April 24, 2023

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-68558-082-7

Location: Venice, Italy

Dates: from April 24, 2023 to April 28, 2023