POLAR: Detecting Multilingual, Multicultural and Multievent Online Polarization

# POLAR: Detecting Multilingual, Multicultural and Multievent Online Polarization
# Based on Naseem et al., SemEval-2026 Task 9
# Paper: https://aclanthology.org/2026.semeval-1.453/
# Task page: https://polar-semeval.github.io/
#
# Given a social-media post (one of 22 languages), judge whether it is
# polarized. The shared task has three subtasks:
#   - Subtask 1: binary polarization detection (polarized / not)
#   - Subtask 2: polarization type (the social dimension involved)
#   - Subtask 3: manifestation categories (how polarization is expressed)
# The organizers characterize polarization as encompassing stereotyping,
# vilification, dehumanization, and intolerance toward social groups; those
# form the manifestation labels below. This showcase collects the binary
# label, the manifestation(s), and the polarization type as free text
# (the official type inventory is defined by the task).

annotation_task_name: "POLAR - Online Polarization"
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: polarized
    description: "Is this post polarized?"
    labels:
      - "Polarized"
      - "Not polarized"
    keyboard_shortcuts:
      "Polarized": "1"
      "Not polarized": "2"

  - annotation_type: multiselect
    name: manifestation
    description: "If polarized, how is it manifested? (Select all that apply)"
    labels:
      - "Stereotyping"
      - "Vilification"
      - "Dehumanization"
      - "Intolerance"
    tooltips:
      "Stereotyping": "Attributing fixed, oversimplified traits to a group"
      "Vilification": "Portraying a group as evil, blameworthy, or threatening"
      "Dehumanization": "Denying a group's humanity (e.g. comparing to animals/objects)"
      "Intolerance": "Rejecting a group's right to exist, speak, or participate"

  - annotation_type: text
    name: polarization_type
    description: "If polarized, what is the target dimension/type (e.g. political, religious, ethnic, gender)? - optional"
    textarea: false
    required: false

annotation_instructions: |
  Read the social-media post (an English gloss is provided for reference).
  First decide whether it is polarized. If it is, select how the
  polarization is manifested (stereotyping, vilification, dehumanization,
  intolerance) and note the target dimension (e.g. political, religious,
  ethnic, gender).

html_layout: |
  <div style="padding: 15px; max-width: 800px; margin: auto;">
    <div style="background: #fff1f2; border: 1px solid #fecdd3; border-radius: 8px; padding: 8px 12px; margin-bottom: 10px;">
      <strong style="color: #9f1239;">Language:</strong> <span>{{language}}</span>
    </div>
    <div style="background: #fef2f2; border: 1px solid #fecaca; border-radius: 8px; padding: 16px; margin-bottom: 8px;">
      <strong style="color: #991b1b;">Post:</strong>
      <p style="font-size: 16px; line-height: 1.7; margin: 8px 0 0 0;">{{text}}</p>
    </div>
    <div style="color: #6b7280; font-size: 13px; font-style: italic;">English gloss (reference): {{gloss}}</div>
  </div>

allow_all_users: true
instances_per_annotator: 50
annotation_per_instance: 3
allow_skip: true