SmartGen: Synthesizing Context-Aware User Behavior Data for Adaptive Smart Home Intelligence
Open Access
- 20 April 2026
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 1751-1762
Abstract
As smart homes become increasingly prevalent, intelligent models are widely used for tasks such as anomaly detection and behavior prediction. These models are typically trained on static datasets, making them brittle to behavioral drift caused by seasonal changes, lifestyle shifts, or evolving routines. However, collecting new behavior data for retraining is often impractical due to its slow pace, high cost, and privacy concerns. In this paper, we propose SmartGen, an LLM-based framework that synthesizes context-aware user behavior data to support continual adaptation of downstream smart home models. SmartGen consists of four key components. First, we design a Time and Semantic-aware Split module to divide long behavior sequences into manageable, semantically coherent subsequences under dual time-span constraints. Second, we propose Semantic-aware Sequence Compression to reduce input length while preserving representative semantics by clustering behavior mapping in latent space. Third, we introduce Graph-guided Sequence Synthesis, which constructs a behavior relationship graph and encodes frequent transitions into prompts, guiding the LLM to generate data aligned with contextual changes while retaining core behavior patterns. Finally, we design a Two-stage Outlier Filter to identify and remove implausible or semantically inconsistent outputs, aiming to improve the factual coherence and behavioral validity of the generated sequences. Experiments on three real-world datasets demonstrate that SmartGen significantly enhances model performance on anomaly detection and behavior prediction tasks under behavioral drift, with anomaly detection improving by 85.43% and behavior prediction by 70.51% on average. The code is available at https://github.com/xzyvoid/SmartGen.Keywords
Funding Information
- National Engineering Laboratory for Technology of Internet Domain Name (KF202512)
- Pengcheng Laboratory (2025QYA001)
- Pengcheng Laboratory (PCL2025A09)
This publication has 31 references indexed in Scilit:
- ProKeR: A Kernel Perspective on Few-Shot Adaptation of Large Vision-Language Models2025
- IoT Anomaly Detection Via Device Interaction Graph2023
- An ML Behavior-Based Security Control for Smart Home Systems2021
- PacketCGAN: Exploratory Study of Class Imbalance for Encrypted Traffic Classification Using CGAN2020
- Flow-based network traffic generation using Generative Adversarial NetworksComputers & Security, 2019
- Self-Attentive Sequential Recommendation2018
- The Synthetic Data Vault2016
- Context-Aware Sequential Recommendation2016
- Synthetic data generation for the internet of things2014
- Isolation Forest2008