MM-Soc: Benchmarking Multimodal Large Language Models in Social Media Platforms

TL;DR

This paper presents MM-Soc, a benchmark for evaluating multimodal large language models' ability to understand social media content, highlighting current limitations and potential improvements through fine-tuning.

Contribution

Introduces MM-Soc, a comprehensive benchmark with new datasets for assessing MLLMs' social media understanding and evaluates multiple models to identify performance gaps.

Findings

MLLMs struggle with social media content in zero-shot settings.

Fine-tuning improves MLLMs' performance on social media tasks.

Significant disparities exist among different MLLMs' capabilities.

Abstract

Social media platforms are hubs for multimodal information exchange, encompassing text, images, and videos, making it challenging for machines to comprehend the information or emotions associated with interactions in online spaces. Multimodal Large Language Models (MLLMs) have emerged as a promising solution to these challenges, yet they struggle to accurately interpret human emotions and complex content such as misinformation. This paper introduces MM-Soc, a comprehensive benchmark designed to evaluate MLLMs' understanding of multimodal social media content. MM-Soc compiles prominent multimodal datasets and incorporates a novel large-scale YouTube tagging dataset, targeting a range of tasks from misinformation detection, hate speech detection, and social context generation. Through our exhaustive evaluation on ten size-variants of four open-source MLLMs, we have identified significant performance disparities, highlighting the need for advancements in models' social understanding capabilities. Our analysis reveals that, in a zero-shot setting, various types of MLLMs generally exhibit difficulties in handling social media tasks. However, MLLMs demonstrate performance improvements post fine-tuning, suggesting potential pathways for improvement. Our code and data are available at https://github.com/claws-lab/MMSoc.git.