New framework SimTrace generates synthetic user data for AI research

Researchers have developed SimTrace, a novel framework designed to generate realistic synthetic user interaction data for online user modeling. This system addresses the scarcity of accessible, fine-grained user trajectories by anonymizing real interactions and simulating web environments. SimTrace aims to provide a privacy-preserving alternative to proprietary logs, enabling advancements in areas like A/B testing, recommender systems, and interface evaluation. The framework has demonstrated strong fidelity and downstream utility, outperforming existing methods and showing comparable performance to models trained on real data for tasks such as purchase prediction and next action prediction. AI

IMPACT Enables development of AI models for user behavior analysis without compromising privacy.

RANK_REASON The cluster describes a new research paper detailing a framework for generating synthetic data. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• SimTrace

• product
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →