V-WSD - Visual Word Sense Disambiguation

# V-WSD - Visual Word Sense Disambiguation
# Based on Raganato et al., SemEval 2023
# Paper: https://aclanthology.org/2023.semeval-1.308/
# Dataset: https://raganato.github.io/vwsd/
#
# Given a target word and a context sentence, annotators select the image
# (from a set of 10 candidates) that best represents the intended meaning
# of the target word in that context. This tests visual grounding of
# word senses across modalities.
#
# Annotation Guidelines:
# 1. Read the context sentence and identify the target word
# 2. Consider the meaning of the target word in the given context
# 3. Review the descriptions of all 10 candidate images
# 4. Select the image that best matches the intended word sense
# 5. If multiple images seem relevant, choose the most specific match

annotation_task_name: "V-WSD - Visual Word Sense Disambiguation"
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: image_selection
    description: "Which image best represents the meaning of the target word in the given context?"
    labels:
      - "Image 1"
      - "Image 2"
      - "Image 3"
      - "Image 4"
      - "Image 5"
      - "Image 6"
      - "Image 7"
      - "Image 8"
      - "Image 9"
      - "Image 10"
    keyboard_shortcuts:
      "Image 1": "1"
      "Image 2": "2"
      "Image 3": "3"
      "Image 4": "4"
      "Image 5": "5"
      "Image 6": "6"
      "Image 7": "7"
      "Image 8": "8"
      "Image 9": "9"
      "Image 10": "0"
    tooltips:
      "Image 1": "Select if Image 1 best represents the target word meaning"
      "Image 2": "Select if Image 2 best represents the target word meaning"
      "Image 3": "Select if Image 3 best represents the target word meaning"
      "Image 4": "Select if Image 4 best represents the target word meaning"
      "Image 5": "Select if Image 5 best represents the target word meaning"
      "Image 6": "Select if Image 6 best represents the target word meaning"
      "Image 7": "Select if Image 7 best represents the target word meaning"
      "Image 8": "Select if Image 8 best represents the target word meaning"
      "Image 9": "Select if Image 9 best represents the target word meaning"
      "Image 10": "Select if Image 10 best represents the target word meaning"

annotation_instructions: |
  You will see a context sentence with a highlighted target word, along with descriptions of 10 candidate images.
  Your task is to select the image that best represents the meaning of the target word as used in the given context.
  Pay attention to the specific sense of the word -- many words have multiple meanings, and the context determines which sense is intended.

html_layout: |
  <div style="padding: 15px; max-width: 800px; margin: auto;">
    <div style="background: #fefce8; border: 1px solid #fde68a; border-radius: 8px; padding: 16px; margin-bottom: 16px;">
      <strong style="color: #a16207;">Target Word:</strong>
      <span style="font-size: 18px; font-weight: bold; color: #854d0e;">{{target_word}}</span>
    </div>
    <div style="background: #f0f9ff; border: 1px solid #bae6fd; border-radius: 8px; padding: 16px; margin-bottom: 16px;">
      <strong style="color: #0369a1;">Context Sentence:</strong>
      <p style="font-size: 16px; line-height: 1.7; margin: 8px 0 0 0;">{{text}}</p>
    </div>
    <div style="background: #f8fafc; border: 1px solid #e2e8f0; border-radius: 8px; padding: 16px;">
      <strong style="color: #475569;">Candidate Image Descriptions:</strong>
      <p style="font-size: 14px; line-height: 1.8; margin: 8px 0 0 0; white-space: pre-line;">{{image_descriptions}}</p>
    </div>
  </div>

allow_all_users: true
instances_per_annotator: 50
annotation_per_instance: 2
allow_skip: true
skip_reason_required: false