Dataset Development 

The dataset was created to study social support in online discourse. It consists of both English- and Spanish-language YouTube comments and is designed as a hierarchical, multi-level annotated dataset covering binary social support detection, support target identification, and group-level categorization

Data Collection 

• Source: YouTube comments
  
  

• Videos: 15 manually selected videos
  
  

• Topics:
  
  

  • National identity
    
    

  • Black Community
    
    

  • Women’s rights
    
    

  • Religion
    
    

  • LGBTQ+
    
    

  • Other socially and emotionally charged topics
    
    

The videos were selected based on:

• Their prominence in public discourse
  
  

• Their potential to elicit supportive or emotional responses
  
  

• High engagement and open comment sections
  
  

Key events include discussions around:

• The Israel–Palestine conflict
  
  

• Black Community–related events
  
  

• LGBTQ+ issues
  
  

• Women’s rights
  
  

• Public reactions to Cristiano Ronaldo following personal life events, which generated a large volume of supportive comments
  
  

Tasks