MAMA-Memeia! Multi-Aspect Multi-Agent Collaboration for Depressive Symptoms Identification in Memes

TL;DR

This paper presents MAMAMemeia, a multi-agent framework leveraging clinical psychology techniques and large language models to detect depressive symptoms in memes, significantly advancing the state-of-the-art in social media mental health analysis.

Contribution

It introduces MAMAMemeia, a novel multi-agent, multi-aspect discussion framework based on Cognitive Analytic Therapy, improving depressive meme detection accuracy and establishing new benchmarks.

Findings

MAMAMemeia outperforms previous methods by 7.55% in macro-F1 score.

The framework is validated on a new resource, RESTOREx, combining LLM-generated and human-annotated explanations.

Achieves state-of-the-art performance on depressive symptom detection in memes.

Abstract

Over the past years, memes have evolved from being exclusively a medium of humorous exchanges to one that allows users to express a range of emotions freely and easily. With the ever-growing utilization of memes in expressing depressive sentiments, we conduct a study on identifying depressive symptoms exhibited by memes shared by users of online social media platforms. We introduce RESTOREx as a vital resource for detecting depressive symptoms in memes on social media through the Large Language Model (LLM) generated and human-annotated explanations. We introduce MAMAMemeia, a collaborative multi-agent multi-aspect discussion framework grounded in the clinical psychology method of Cognitive Analytic Therapy (CAT) Competencies. MAMAMemeia improves upon the current state-of-the-art by 7.55% in macro-F1 and is established as the new benchmark compared to over 30 methods.