Text2Freq: Learning Series Patterns from Text via Frequency Domain

TL;DR

Text2Freq introduces a novel frequency domain approach to integrate textual descriptions with time series data, enhancing forecasting accuracy by aligning text with low-frequency components, and demonstrates state-of-the-art results on stock datasets.

Contribution

The paper presents a new cross-modality model that effectively aligns text and time series data in the frequency domain, addressing the modality gap and improving forecasting performance.

Findings

Achieves state-of-the-art forecasting accuracy on stock datasets.

Effectively aligns textual information with low-frequency components of time series.

Encourages future research in multimodal time series analysis.

Abstract

Traditional time series forecasting models mainly rely on historical numeric values to predict future outcomes.While these models have shown promising results, they often overlook the rich information available in other modalities, such as textual descriptions of special events, which can provide crucial insights into future dynamics.However, research that jointly incorporates text in time series forecasting remains relatively underexplored compared to other cross-modality work. Additionally, the modality gap between time series data and textual information poses a challenge for multimodal learning. To address this task, we propose Text2Freq, a cross-modality model that integrates text and time series data via the frequency domain. Specifically, our approach aligns textual information to the low-frequency components of time series data, establishing more effective and interpretable alignments between these two modalities. Our experiments on paired datasets of real-world stock prices and synthetic texts show that Text2Freq achieves state-of-the-art performance, with its adaptable architecture encouraging future research in this field.