iSarcasmEval: Intended Sarcasm Detection
SemEval-2022 Task 6 (iSarcasmEval): detecting intended sarcasm in English and Arabic. Unlike prior datasets labeled by third parties, iSarcasmEval texts are labeled by their own authors for intended sarcasm. Annotators decide whether a text is sarcastic and, if so, which category of ironic speech it expresses.
Configuration Fileconfig.yaml
This Potato config reproduces the annotation task. Save it as config.yaml and run potato start config.yaml to try it.
# iSarcasmEval: Intended Sarcasm Detection
# Based on Abu Farha, Oprea, Wilson, and Magdy, SemEval-2022 Task 6
# Paper: https://aclanthology.org/2022.semeval-1.111/
# Dataset: https://github.com/iabufarha/iSarcasmEval
#
# iSarcasmEval is the first shared task to focus on INTENDED sarcasm: each
# text is labeled by its own author, not by third-party annotators. The
# shared task defines three subtasks:
# - SubTask A: is the text sarcastic? (binary)
# - SubTask B: which category of ironic speech does it express?
# - SubTask C: given a sarcastic text and its non-sarcastic rephrase,
# tell which one is the sarcastic one (pairwise)
# This showcase models SubTask A (binary) and SubTask B (category), and
# collects an optional non-sarcastic rephrase in the spirit of SubTask C.
#
# SubTask B ironic-speech categories (Leggitt & Gibbs):
# - Sarcasm: contradicts the state of affairs and is critical toward someone
# - Irony: contradicts the state of affairs but is not necessarily critical
# - Satire: ridicules a subject through humor, irony, or exaggeration
# - Understatement: undermines the magnitude of a state of affairs
# - Overstatement: exaggerates the magnitude of a state of affairs
# - Rhetorical Question: a question whose answer is not expected
annotation_task_name: "iSarcasmEval - Intended Sarcasm Detection"
task_dir: "."
data_files:
- sample-data.json
item_properties:
id_key: "id"
text_key: "text"
output_annotation_dir: "annotation_output/"
output_annotation_format: "json"
port: 8000
server_name: localhost
annotation_schemes:
- annotation_type: radio
name: is_sarcastic
description: "Is this text sarcastic (as intended by its author)?"
labels:
- "Sarcastic"
- "Not Sarcastic"
keyboard_shortcuts:
"Sarcastic": "1"
"Not Sarcastic": "2"
tooltips:
"Sarcastic": "The author intended the text to be sarcastic or ironic"
"Not Sarcastic": "The text is meant literally"
- annotation_type: multiselect
name: sarcasm_category
description: "If sarcastic, which category of ironic speech applies? (Select all that apply)"
labels:
- "Sarcasm"
- "Irony"
- "Satire"
- "Understatement"
- "Overstatement"
- "Rhetorical Question"
tooltips:
"Sarcasm": "Contradicts the state of affairs and is critical toward someone/something"
"Irony": "Contradicts the state of affairs but is not necessarily critical"
"Satire": "Ridicules a subject through humor, irony, or exaggeration"
"Understatement": "Undermines the magnitude of a state of affairs"
"Overstatement": "Exaggerates the magnitude of a state of affairs"
"Rhetorical Question": "A question posed without expecting an answer"
- annotation_type: text
name: rephrase
description: "If sarcastic, provide a non-sarcastic rephrase that conveys the intended meaning (optional)"
textarea: true
required: false
placeholder: "Non-sarcastic rephrase..."
annotation_instructions: |
You will see a short social-media text. First decide whether it is
sarcastic as intended by its author. If it is, select the category or
categories of ironic speech it expresses, and optionally write a
non-sarcastic rephrase that conveys the same intended meaning.
html_layout: |
<div style="padding: 15px; max-width: 800px; margin: auto;">
<div style="background: #fff7ed; border: 1px solid #fed7aa; border-radius: 8px; padding: 16px; margin-bottom: 12px;">
<strong style="color: #9a3412;">Text:</strong>
<p style="font-size: 17px; line-height: 1.7; margin: 8px 0 0 0;">{{text}}</p>
</div>
</div>
allow_all_users: true
instances_per_annotator: 50
annotation_per_instance: 3
allow_skip: true
Sample Datasample-data.json
[
{
"id": "isc_001",
"text": "Oh great, another Monday. Just what I was hoping for all weekend."
},
{
"id": "isc_002",
"text": "The train was delayed by an hour this morning due to signal problems."
}
]
// ... and 8 more itemsGet This Design
Clone or download from the repository
Quick start:
git clone https://github.com/davidjurgens/potato-showcase.git cd potato-showcase/semeval/2022/task06-isarcasmeval potato start config.yaml
Dataset & paper
Abu Farha et al., SemEval 2022
Citation (BibTeX)
@inproceedings{abu-farha-etal-2022-semeval,
title = "{S}em{E}val-2022 Task 6: i{S}arcasm{E}val, Intended Sarcasm Detection in {E}nglish and {A}rabic",
author = "Abu Farha, Ibrahim and Oprea, Silviu Vlad and Wilson, Steven and Magdy, Walid",
booktitle = "Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022)",
month = jul,
year = "2022",
address = "Seattle, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2022.semeval-1.111",
doi = "10.18653/v1/2022.semeval-1.111",
pages = "802--814"
}Details
Annotation Types
Domain
Use Cases
Tags
Found an issue or want to improve this design?
Open an IssueRelated Designs
POLAR: Detecting Multilingual, Multicultural and Multievent Online Polarization
SemEval-2026 Task 9 (POLAR): detecting online polarization across 22 languages, many cultures, and multiple real-world events. Annotators judge whether a social-media post is polarized and, if so, identify the type of polarization and how it is rhetorically manifested (e.g. stereotyping, vilification, dehumanization, intolerance). Polarization encompasses stereotyping, vilification, dehumanization, and intolerance directed at social groups.
FLUTE: Figurative Language Understanding through Textual Explanations
Figurative language understanding via NLI. Annotators classify figurative sentences (sarcasm, simile, metaphor, idiom) and provide textual explanations of the figurative meaning. The task combines natural language inference with fine-grained figurative language type classification.
HatEval - Multilingual Detection of Hate Speech Against Immigrants and Women
Detection and classification of hate speech targeting immigrants and women on Twitter, including fine-grained annotation of target range (individual vs. generic) and aggressiveness. Based on SemEval-2019 Task 5 (HatEval).