🤖 AI Summary
This work addresses the challenge of accurately mapping natural language descriptions to temporal intervals in multivariate time series for semantic event detection under scarce annotation. To this end, the authors propose a knowledge-guided neuro-symbolic framework that introduces, for the first time, an Event Logic Tree (ELT) to structurally align linguistic semantics with temporal logic over time series. Integrating a vision-language model (VLM) agent, the framework iteratively instantiates signal visualization primitives under ELT constraints, enabling zero-shot and interpretable event detection. Evaluated on real-world data and a newly curated expert-annotated benchmark, the method substantially outperforms both supervised fine-tuning approaches and existing large language model (LLM)/VLM zero-shot baselines. Human evaluations further confirm its high detection accuracy and trustworthy interpretability.
📝 Abstract
Time Series Event Detection (TSED) has long been an important task with critical applications across many high-stakes domains. Unlike statistical anomalies, events are defined by semantics with complex internal structures, which are difficult to learn inductively from scarce labeled data in real-world settings. In light of this, we introduce Knowledge-Guided TSED, a new setting where a model is given a natural-language event description and must ground it to intervals in multivariate signals with little or no training data. To tackle this challenge, we introduce Event Logic Tree (ELT), a novel knowledge representation framework to bridge linguistic descriptions and physical time series data via modeling the intrinsic temporal-logic structures of events. Based on ELT, we present a neuro-symbolic VLM agent framework that iteratively instantiates primitives from signal visualizations and composes them under ELT constraints, producing both detected intervals and faithful explanations in the form of instantiated trees. To validate the effectiveness of our approach, we release a benchmark based on real-world time series data with expert knowledge and annotations. Experiments and human evaluation demonstrate the superiority of our method compared to supervised fine-tuning baselines and existing zero-shot time series reasoning frameworks based on LLMs/VLMs. We also show that ELT is critical in mitigating VLMs' inherent hallucination in matching signal morphology with event semantics.