# Exploring meaning in life from social network content in the sleep scenario

## TL;DR

This study uses social media data and AI to explore how people express meaning in life, especially around sleep, and identifies factors that may affect mental health.

## Contribution

The novel contribution is combining deep learning, semantic graphs, and social media data to analyze meaning in life and its psychological risk factors.

## Key findings

- Deep learning models were developed to assess components of meaning in life from social media posts.
- Semantic dependency graphs identified factors associated with meaning in life and their interrelations.
- Eight subnetworks revealed differences in meaning in life at individual and group levels.

## Abstract

The exploration of life’s meaning has been a key topic across disciplines, and artificial intelligence is now beginning to investigate it.

This study leveraged social media to assess meaning in life (MIL) and its associated factors at individual and group levels. We compiled a diverse dataset consisting of microblog posts (N = 7,588,597) and responses from user surveys (N = 448), annotated using a combination of self-assessment, expert opinions, and ChatGPT-generated insights. Our methodology examined MIL in three ways: (1) developing deep learning models to assess MIL components, (2) applying semantic dependency graph algorithms to identify MIL associated factors, and (3) constructing eight subnetworks to analyze factors, their interrelations, and MIL differences.

We validated these methods and bridged two foundational MIL theories, highlighting their interconnections.

By identifying psychological risk factors, our work may provide clues to mental health issues and inform possible intervention.