MM-SOC: Benchmarking Multimodal Large Language Models in Social Media Platforms
Model architecture and key componentsAbstract
Social media platforms are hubs for multimodal information exchange, encompassing text, images, and videos, making it challenging for machines to comprehensively understand the information. Multimodal Large Language Models (MLLMs) have shown promise in addressing these challenges, yet they struggle with accurately interpreting the intertwined multimodal cues in social media content. We introduce MM-Soc, a comprehensive benchmark designed to evaluate MLLMs’ understanding of multimodal social media content.
Type
Publication
ACL (Findings) 2024
Abstract
Social media platforms are hubs for multimodal information exchange, encompassing text, images, and videos, making it challenging for machines to comprehensively understand the information. Multimodal Large Language Models (MLLMs) have shown promise in addressing these challenges, yet they struggle with accurately interpreting the intertwined multimodal cues in social media content. We introduce MM-Soc, a comprehensive benchmark designed to evaluate MLLMs’ understanding of multimodal social media content.
Keywords
Multimodal Learning, Large Language Models, Social Media Analysis, Benchmarking
