Combining time-series and textual data for taxi demand prediction in event areas: a deep learning approach

TL;DR

This paper introduces deep learning models that combine time-series and textual data to improve taxi demand forecasting in event areas, demonstrating significant error reduction using real-world data.

Contribution

It presents novel architectures that integrate textual explanations with temporal data for demand prediction, a previously underexplored approach in this context.

Findings

Models significantly reduce forecasting error.

Combining text and time-series improves accuracy.

Deep learning architectures effectively fuse multimodal data.

Abstract

Accurate time-series forecasting is vital for numerous areas of application such as transportation, energy, finance, economics, etc. However, while modern techniques are able to explore large sets of temporal data to build forecasting models, they typically neglect valuable information that is often available under the form of unstructured text. Although this data is in a radically different format, it often contains contextual explanations for many of the patterns that are observed in the temporal data. In this paper, we propose two deep learning architectures that leverage word embeddings, convolutional layers and attention mechanisms for combining text information with time-series data. We apply these approaches for the problem of taxi demand forecasting in event areas. Using publicly available taxi data from New York, we empirically show that by fusing these two complementary cross-modal sources of information, the proposed models are able to significantly reduce the error in the forecasts.