Time-MQA: Time Series Multi-Task Question Answering with Context Enhancement

2025-02-26

AI Summary

Existing time-series methods typically target single tasks (e.g., forecasting or anomaly detection) and lack a unified natural-language interface supporting numerical analysis and open-ended reasoning. This paper introduces the first multi-task time-series question answering (TSQA) paradigm, enabling natural-language interaction across diverse long- and short-horizon time-series scenarios—from environmental monitoring to traffic management. Our contributions are threefold: (1) We release TSQA, the first large-scale time-series QA dataset comprising ~200K question-answer pairs; (2) We conduct continual pretraining of Mistral-7B, Llama-3-8B, and Qwen-2.5-7B to enhance temporal semantic understanding and cross-task reasoning; (3) We propose a context-augmented time-series–language alignment mechanism. Experiments demonstrate significant improvements in temporal reasoning accuracy, cross-task generalization, and natural-language interaction quality. All data, models, code, and evaluation results are publicly available.

Technology Category

• Natural Language Processing: Question Answering
• Machine Learning: Time-Series/Data Streams
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web question answering
• Web Mining and Content Analysis: Large pretrained models with web data
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Time series data are foundational in finance, healthcare, and energy domains. However, most existing methods and datasets remain focused on a narrow spectrum of tasks, such as forecasting or anomaly detection. To bridge this gap, we introduce Time Series Multi-Task Question Answering (Time-MQA), a unified framework that enables natural language queries across multiple time series tasks - numerical analytical tasks and open-ended question answering with reasoning. Central to Time-MQA is the TSQA dataset, a large-scale dataset containing $sim$200k question-answer pairs derived from diverse time series spanning environment, traffic, etc. This comprehensive resource covers various time series lengths and promotes robust model development. We further demonstrate how continually pre-training large language models (Mistral 7B, Llama-3 8B, and Qwen-2.5 7B) on the TSQA dataset enhanced time series reasoning capabilities, moving beyond mere numeric tasks and enabling more advanced and intuitive interactions with temporal data. The complete TSQA dataset, models, executable codes, user study questionnaires for evaluation, and results have all been open-sourced.

Problem

• Unified framework for multi-task time series question answering.
• Enhances time series reasoning with large language models.
• Introduces TSQA dataset for robust model development.

Innovation

• Unified framework for multi-task time series queries
• Large-scale TSQA dataset with 200k question-answer pairs
• Enhanced time series reasoning via pre-trained language models