SigTime: Learning and Visually Explaining Time Series Signatures

2025-12-12

AI Summary

Time-series pattern understanding faces challenges including high computational complexity, poor interpretability, and difficulty in modeling meaningful temporal structures. To address these, we propose SigTime—a novel dual-Transformer architecture that jointly learns shapelet-based local morphological patterns and statistical features in an end-to-end manner, generating discriminative and semantically interpretable time-series signatures. SigTime introduces a differentiable shapelet learning mechanism and a multi-scale temporal representation fusion strategy. Furthermore, we develop an interactive, multi-view visualization system built on D3/React for exploratory analysis. Evaluated on eight public benchmarks and one clinical dataset, SigTime achieves state-of-the-art classification performance. Clinical expert evaluations confirm its effectiveness in enhancing pattern insight efficiency and decision-making trustworthiness—particularly in ECG diagnosis and preterm birth risk assessment.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning
• Computer Vision: Multi-modal Vision

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding

Abstract

Understanding and distinguishing temporal patterns in time series data is essential for scientific discovery and decision-making. For example, in biomedical research, uncovering meaningful patterns in physiological signals can improve diagnosis, risk assessment, and patient outcomes. However, existing methods for time series pattern discovery face major challenges, including high computational complexity, limited interpretability, and difficulty in capturing meaningful temporal structures. To address these gaps, we introduce a novel learning framework that jointly trains two Transformer models using complementary time series representations: shapelet-based representations to capture localized temporal structures and traditional feature engineering to encode statistical properties. The learned shapelets serve as interpretable signatures that differentiate time series across classification labels. Additionally, we develop a visual analytics system -- SigTIme -- with coordinated views to facilitate exploration of time series signatures from multiple perspectives, aiding in useful insights generation. We quantitatively evaluate our learning framework on eight publicly available datasets and one proprietary clinical dataset. Additionally, we demonstrate the effectiveness of our system through two usage scenarios along with the domain experts: one involving public ECG data and the other focused on preterm labor analysis.

Problem

• Addresses high computational complexity in time series pattern discovery
• Improves interpretability of learned temporal patterns and signatures
• Captures meaningful temporal structures using complementary representations

Innovation

• Transformer models with shapelet-based and statistical representations
• Interpretable shapelets as signatures for time series classification
• Visual analytics system for multi-perspective signature exploration