skip to main content
introduction
Free access

ACM TIST Special Issue on Deep Learning for Spatio-Temporal Data: Part 1

Published: 16 December 2021 Publication History

1 Introduction

With the quick development of different position techniques, such as Global Position System (GPS), mobile devices and remote sensing, various spatio-temporal data have been generated nowadays. It is of great importance for many practical applications, including human mobility mining, urban planning, health care, and public safety, to mine valuable knowledges from spatio-temporal data. Recently, deep learning has achieved considerable success in many domains, such as computer vision, natural language processing, and medical analysis. It also gets remarkable performance gains in various spatio-temporal data mining (STDM) tasks, including crowd flow prediction, origin-destination (OD) prediction, and health care. The aim of this special issue is to show the latest research findings and engineering experiences in developing and applying deep learning techniques for spatio-temporal data mining tasks and applications from the perspective of academia and industry.
The objective of this special issue is to highlight leading work in spatio-temporal data mining with deep learning techniques, to identify challenges, and to explore future topics in this area. The call for papers of this special issue was perceived very positively with 78 regular submissions, and 22 papers were finally accepted through a rigorous review process. The acceptance rate was therefore 28%. The accepted articles are divided into two parts for publication, and each part includes 11 articles. A brief introduction to the articles of part 1 is presented as follows.
In the article titled “Temporal Hierarchical Graph Attention Network for Traffic Prediction,” Huang et al. propose a novel Temporal Hierarchical Graph Attention Network (TH-GAT) to consider the hierarchical regional structure of the road network and spatio-temporal dependencies between nodes and regions. Efficient experiments conducted on two real-world traffic datasets confirm the superiority of the proposed TH-GAT model.
In the article titled “POLLA: Enhancing the Local Structure Awareness in Long Sequence Spatial-temporal Modeling,” Zhou et al. study the problem of spatio-temporal modeling on long sequences which is practically important in many real-world applications. To consider the underlying structure information of long sequence, they propose a Proximity-aware Long Sequence Learning (POLLA) framework and apply it to the spatio-temporal prediction tasks. Experiments results on five large-scale datasets demonstrate the superiority and effectiveness of POLLA.
In the article titled “A Dynamic Convolutional Neural Network Based Shared-bike Demand Forecasting Model,” Qiao et al. develop a new Shared-bike Demand Forecasting Model based on dynamic convolutional neural networks named SDF to predict the demand of shared bikes. SDF has the ability to choose the most relevant weather features from real weather data by taking into account the states of stations from historical data. The experimental results of SDF are more accurate and efficient than classical machine learning models.
In the article titled “TWIST-GAN: Toward Wavelet Transform and Transferred GAN for Spatio-Temporal Single Image Super Resolution,” Dharejo et al. study the Single Image Super-resolution (SISR) of remotely sensed images with low spatial resolution. To consider texture-feature representation and high-frequency information, the authors propose a frequency domain-based spatio-temporal remote sensing single image super-resolution technique named TWIST-GAN. Qualitative and quantitative evaluations show the proposed framework is superior to state-of-the-art approaches. The GPUS memory usage of TWIST-GAN is much lower than other methods.
In the article titled “TARA-Net: A Fusion Network for Detecting Takeaway Rider Accidents,” He et al. have found that rider traffic accidents raise financial cost and social traffic burden. They study the problem of detecting the takeaway rider accidents based on food delivery trajectory data, which was not studied before. A TARA-Net is proposed to jointly model the heterogeneous and spatio-temporal sequence data for takeaway rider accident detection. Extensive experiments demonstrate that their proposed TARA-Net makes sense.
In the article titled “PARP: A Parallel Traffic Condition Driven Route Planning Model on Dynamic Road Networks,” this work concentrates on route planning which is essential to many location-based services. The authors design a practical route planning strategy PARP by employing a GCN model and a dual-level path index to embed the future traffic condition and reduce the response time, respectively. They also conduct extensive evaluations to confirm the effectiveness and superiority of PARP.
In the article titled “Route Optimization via Environment-aware Deep Network and Reinforcement Learning,” Guo et al. develop a mobile sequential recommendation system for vehicle route optimization in urban areas. A reinforcement-learning framework is proposed to integrate a self-check mechanism and a deep neural network for customer pick-up point monitoring. Extensive experiment results show the performance of the proposed method is excellent consistently from hourly to weekly measures.
In the article titled “TAML: A Traffic-aware Multi-task Learning Model for Estimating Travel Time,” Xu et al. make their effort to estimate travel time, which is important in broad real-world applications. To capture the mutual influence of travel time and traffic condition, they propose an improved multi-task traffic-aware model called TAML consisting of a travel time estimator and a traffic estimator. Extensive experiments on two real trajectory datasets show that TAML is effective.
In the article titled “Spatial Variability Aware Deep Neural Networks (SVANN): A General Approach,” Gupta et al. explore spatial variability feature of various geographical phenomena and propose a novel Spatial Variability Aware Deep Neural Network ASVANN-E that is a more flexible extension of their previous work SVANN. The architecture of SVANN-E does not follow a spatial-one-size-fits-all (OSFA) approach, and it varies across geographical locations in order to quantify spatial variability. The experimental results demonstrate that SVANN-E outperforms state-of-the-art methods.
In the article titled “Similar Trajectory Search with Spatio-Temporal Deep Representation Learning,” similar trajectory search is studied which is crucial and can facilitate many downstream spatial data analytic applications. To better consider spatial and temporal correlations, the authors propose a spatio-temporal deep representation learning-based approach to search similar trajectory. Experiments show the proposed method achieves significant performance improvement over existing approaches.
In the article titled “Predicting Human Mobility with Reinforcement-learning-based Long-term Periodicity Modeling,” Tao et al. propose MoveNet and RLMoveNet for human mobility prediction. MoveNet is a self-attention-based sequential model aiming at predicting each user’s next destination based on their historical trajectories. Then, they add the reinforcement learning layer and further propose RLMoveNet which considers human mobility prediction as a reinforcement learning problem. The performance of MoveNet and RLMoveNet is more outstanding than the state-of-the-art mobility prediction methods.
Part 1 of this special issue covers the topics of spatio-temporal data modeling, representation learning, and various important applications including traffic prediction, route planning, and human trajectory prediction with designed deep learning models. By no means can one special issue scope all the exciting research in the field of spatio-temporal data mining with deep learning techniques. In addition, more challenges are emerging with more real-time data sensing, more spatio-temporal data related applications, and more powerful processing capacities. We thank all the authors for their contributions. We appreciate the valuable comments from our reviewers, and we are grateful to the editorial board of ACM TIST for their help in coordinating the publication process. We hope that this collection will motivate further research in the exciting research area of deep learning-based spatio-temporal data mining.

Cited By

View all
  • (2024)Generative Adversarial Network Applications in Industry 4.0: A ReviewInternational Journal of Computer Vision10.1007/s11263-023-01966-9132:6(2195-2254)Online publication date: 12-Jan-2024
  • (2024)An improved normal wiggly hesitant fuzzy FMEA model and its application to risk assessment of electric bus systemsApplied Intelligence10.1007/s10489-024-05458-254:8(6213-6237)Online publication date: 8-May-2024
  • (2023)A Spatial and Adversarial Representation Learning Approach for Land Use Classification with POIsACM Transactions on Intelligent Systems and Technology10.1145/362782414:6(1-25)Online publication date: 14-Nov-2023
  • Show More Cited By

Index Terms

  1. ACM TIST Special Issue on Deep Learning for Spatio-Temporal Data: Part 1
        Index terms have been assigned to the content through auto-classification.

        Recommendations

        Comments

        Information & Contributors

        Information

        Published In

        cover image ACM Transactions on Intelligent Systems and Technology
        ACM Transactions on Intelligent Systems and Technology  Volume 12, Issue 6
        December 2021
        356 pages
        ISSN:2157-6904
        EISSN:2157-6912
        DOI:10.1145/3501281
        • Editor:
        • Huan Liu
        Issue’s Table of Contents

        Publisher

        Association for Computing Machinery

        New York, NY, United States

        Publication History

        Published: 16 December 2021
        Published in TIST Volume 12, Issue 6

        Permissions

        Request permissions for this article.

        Check for updates

        Qualifiers

        • Introduction
        • Opinion
        • Refereed

        Contributors

        Other Metrics

        Bibliometrics & Citations

        Bibliometrics

        Article Metrics

        • Downloads (Last 12 months)230
        • Downloads (Last 6 weeks)37
        Reflects downloads up to 15 Feb 2025

        Other Metrics

        Citations

        Cited By

        View all
        • (2024)Generative Adversarial Network Applications in Industry 4.0: A ReviewInternational Journal of Computer Vision10.1007/s11263-023-01966-9132:6(2195-2254)Online publication date: 12-Jan-2024
        • (2024)An improved normal wiggly hesitant fuzzy FMEA model and its application to risk assessment of electric bus systemsApplied Intelligence10.1007/s10489-024-05458-254:8(6213-6237)Online publication date: 8-May-2024
        • (2023)A Spatial and Adversarial Representation Learning Approach for Land Use Classification with POIsACM Transactions on Intelligent Systems and Technology10.1145/362782414:6(1-25)Online publication date: 14-Nov-2023
        • (2023)Evaluation Framework for Electric Vehicle Security Risk AssessmentIEEE Transactions on Intelligent Transportation Systems10.1109/TITS.2023.330766025:1(33-56)Online publication date: 11-Sep-2023
        • (2023)User re-identification via human mobility trajectories with siamese transformer networksApplied Intelligence10.1007/s10489-023-05234-854:1(815-834)Online publication date: 20-Dec-2023
        • (2023)Panini: a transformer-based grammatical error correction method for BanglaNeural Computing and Applications10.1007/s00521-023-09211-736:7(3463-3477)Online publication date: 4-Dec-2023
        • (undefined)Research on the Application of Neural Network Classification Model in English Grammar Error CorrectionACM Transactions on Asian and Low-Resource Language Information Processing10.1145/3596492

        View Options

        View options

        PDF

        View or Download as a PDF file.

        PDF

        eReader

        View online with eReader.

        eReader

        HTML Format

        View this article in HTML Format.

        HTML Format

        Login options

        Full Access

        Figures

        Tables

        Media

        Share

        Share

        Share this Publication link

        Share on social media