When LLM Meets Hypergraph: A Sociological Analysis on Personality via Online Social Networks

TL;DR

This work addresses how to analyze personality in online social networks from a sociological perspective by integrating environment-aware hypergraph modeling with prompting-based LLM semantic enhancement. It introduces a novel dataset collected from PersonalityCafe that includes user profiles, personality labels, and detailed social environments, enabling environment-rich analysis. The framework models three hyperedge types to capture social contexts, uses LLM-generated narratives to densify fragmented user data, and employs a hypergraph neural network with skip connections and focal loss to predict MBTI and Enneagram traits. Experimental results show the proposed method outperforms dyadic GNNs and LLM-only baselines, demonstrating the value of environment-aware representations and semantic augmentation for nuanced personality inference in digital social systems.

Abstract

Individual personalities significantly influence our perceptions, decisions, and social interactions, which is particularly crucial for gaining insights into human behavior patterns in online social network analysis. Many psychological studies have observed that personalities are strongly reflected in their social behaviors and social environments. In light of these problems, this paper proposes a sociological analysis framework for one's personality in an environment-based view instead of individual-level data mining. Specifically, to comprehensively understand an individual's behavior from low-quality records, we leverage the powerful associative ability of LLMs by designing an effective prompt. In this way, LLMs can integrate various scattered information with their external knowledge to generate higher-quality profiles, which can significantly improve the personality analysis performance. To explore the interactive mechanism behind the users and their online environments, we design an effective hypergraph neural network where the hypergraph nodes are users and the hyperedges in the hypergraph are social environments. We offer a useful dataset with user profile data, personality traits, and several detected environments from the real-world social platform. To the best of our knowledge, this is the first network-based dataset containing both hypergraph structure and social information, which could push forward future research in this area further. By employing the framework on this dataset, we can effectively capture the nuances of individual personalities and their online behaviors, leading to a deeper understanding of human interactions in the digital world.

Figures (9)

• Figure 1: Traditional Approach v.s. Sociological Approach
• Figure 2: The Framework of Our Sociological Personality Analysis. An agent reads the fragmented data and injects the raw information into a carefully designed prompt, which will be used to interact with an LLM. The enhanced content and three kinds of social environments are sent into a hypergraph neural network for sociological personality analysis.
• Figure 3: The Process of Enhancing User Semantic Description with Fragmented Information.
• Figure 4: Personality distribution (MBTI)
• Figure 5: Data Distribution Grouped by Personality