CometNet: Contextual Motif-guided Long-term Time Series Forecasting

2025-11-11

AI Summary

Long-term time series forecasting is hindered by narrow model receptive fields; mainstream Transformer- and MLP-based approaches rely on fixed-length historical windows, limiting their capacity to capture long-range temporal dependencies—while naively extending the window incurs prohibitive computational overhead and noise amplification. To address this, we propose a context motif modeling paradigm that adaptively identifies recurrent, discriminative temporal patterns from historical sequences and leverages them via a motif-guided prediction mechanism to enable global, cross-window dependency modeling. Crucially, our method introduces no additional parameters or computational cost and integrates seamlessly into existing Transformer or MLP architectures. Evaluated on eight real-world datasets, it consistently outperforms state-of-the-art methods, delivering stable and substantial improvements—particularly in long-horizon forecasting (≥96 steps).

Technology Category

• Machine Learning: Time-Series/Data Streams
• Computer Vision: Multi-modal Vision
• Natural Language Processing: (Large) Language Models

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search

Abstract

Long-term Time Series Forecasting is crucial across numerous critical domains, yet its accuracy remains fundamentally constrained by the receptive field bottleneck in existing models. Mainstream Transformer- and Multi-layer Perceptron (MLP)-based methods mainly rely on finite look-back windows, limiting their ability to model long-term dependencies and hurting forecasting performance. Naively extending the look-back window proves ineffective, as it not only introduces prohibitive computational complexity, but also drowns vital long-term dependencies in historical noise. To address these challenges, we propose CometNet, a novel Contextual Motif-guided Long-term Time Series Forecasting framework. CometNet first introduces a Contextual Motif Extraction module that identifies recurrent, dominant contextual motifs from complex historical sequences, providing extensive temporal dependencies far exceeding limited look-back windows; Subsequently, a Motif-guided Forecasting module is proposed, which integrates the extracted dominant motifs into forecasting. By dynamically mapping the look-back window to its relevant motifs, CometNet effectively harnesses their contextual information to strengthen long-term forecasting capability. Extensive experimental results on eight real-world datasets have demonstrated that CometNet significantly outperforms current state-of-the-art (SOTA) methods, particularly on extended forecast horizons.

Problem

• Overcoming receptive field bottleneck in long-term time series forecasting
• Addressing limitations of finite look-back windows in dependency modeling
• Resolving computational complexity and noise issues in extended historical data

Innovation

• Identifies recurrent contextual motifs from historical sequences
• Integrates extracted dominant motifs into forecasting process
• Dynamically maps look-back windows to relevant motifs