Semantic-aware Graph-guided Behavior Sequences Generation with Large Language Models for Smart Homes

TL;DR

SmartGen is a novel LLM-based framework that synthesizes context-aware, semantically coherent user behavior data to enable smart home models to adapt continually to behavioral changes, improving anomaly detection and prediction.

Contribution

The paper introduces SmartGen, a comprehensive framework with novel modules for sequence splitting, compression, graph-guided synthesis, and outlier filtering to generate realistic behavior sequences for smart home models.

Findings

Significant improvement in anomaly detection accuracy by 85.43%.

Behavior prediction accuracy increased by 70.51%.

Effective handling of behavioral drift in real-world datasets.

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/horizonsinzqs/SmartGen.