Large language model as user daily behavior data generator: balancing population diversity and individual personality

2025-05-23

AI Summary

Human daily behavior prediction faces challenges including complex behavioral patterns, strong short-term volatility, and scarcity of privacy-sensitive real-world data. To address these, we propose BehaviorGen—the first synthetic data generation framework that systematically leverages large language models (LLMs) for general human behavior modeling. BehaviorGen integrates user profiling with event-driven behavioral simulation, jointly optimizing individual personality specificity and population-level diversity. It generates high-fidelity, privacy-preserving synthetic behavioral sequences to support three downstream use cases: pretraining augmentation, fine-tuning replacement, and fine-tuning augmentation. Evaluated on human mobility and smartphone usage prediction tasks, BehaviorGen achieves up to 18.9% improvement in prediction accuracy over state-of-the-art synthetic data methods, demonstrating superior fidelity and utility.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Predicting human daily behavior is challenging due to the complexity of routine patterns and short-term fluctuations. While data-driven models have improved behavior prediction by leveraging empirical data from various platforms and devices, the reliance on sensitive, large-scale user data raises privacy concerns and limits data availability. Synthetic data generation has emerged as a promising solution, though existing methods are often limited to specific applications. In this work, we introduce BehaviorGen, a framework that uses large language models (LLMs) to generate high-quality synthetic behavior data. By simulating user behavior based on profiles and real events, BehaviorGen supports data augmentation and replacement in behavior prediction models. We evaluate its performance in scenarios such as pertaining augmentation, fine-tuning replacement, and fine-tuning augmentation, achieving significant improvements in human mobility and smartphone usage predictions, with gains of up to 18.9%. Our results demonstrate the potential of BehaviorGen to enhance user behavior modeling through flexible and privacy-preserving synthetic data generation.

Problem

• Predicting human daily behavior is complex and variable
• Privacy concerns limit access to large-scale user data
• Generating synthetic behavior data preserves privacy and enhances prediction

Innovation

• LLM-based synthetic behavior data generation
• Balancing diversity and personality in data
• Privacy-preserving user behavior modeling