Findings of the CONSTRAINT 2022 Shared Task on Detecting the Hero, the Villain, and the Victim in Memes

We present the findings of the shared task at the CONSTRAINT 2022 workshop on "Hero, Villain, and Victim: Dissecting Harmful Memes for Semantic Role Labeling of Entities." The task aims to delve deeper into meme comprehension by deciphering the connotations behind the entities present in a meme. In more nuanced terms, the shared task focuses on determining the victimizing, glorifying, and vilifying intentions embedded in meme entities to explicate their connotations. To this end, we curate HVVMemes, a novel meme dataset of about 7,000 memes spanning the domains of COVID-19 and US Politics, each containing entities and their associated roles: hero, villain, victim, or other. The shared task attracted 105 registered participants, but eventually only nine of them made official submissions. The most successful systems used ensembles combining textual and multimodal models, with the best system achieving an F1-score of 58.67.