MTrajRec: Map-Constrained Trajectory Recovery via Seq2Seq Multi-task Learning

With the increasing adoption of GPS modules, there are a wide range of urban applications based on trajectory data analysis, such as vehicle navigation, travel time estimation, and driver behavior analysis. The effectiveness of urban applications relies greatly on the high sampling rates of trajectories precisely matched to the map. However, a large number of trajectories are collected under a low sampling rate in real-world practice, due to certain communication loss and energy constraints. To enhance the trajectory data and support the urban applications more effectively, many trajectory recovery methods are proposed to infer the trajectories in free space. In addition, the recovered trajectory still needs to be mapped to the road network, before it can be used in the applications. However, the two-stage pipeline, which first infers high-sampling-rate trajectories and then performs the map matching, is inaccurate and inefficient. In this paper, we propose a Map-constrained Trajectory Recovery framework, MTrajRec, to recover the fine-grained points in trajectories and map match them on the road network in an end-to-end manner. MTrajRec implements a multi-task sequence-to-sequence learning architecture to predict road segment and moving ratio simultaneously. Constraint mask, attention mechanism, and attribute module are proposed to overcome the limits of coarse grid representation and improve the performance. Extensive experiments based on large-scale real-world trajectory data confirm the effectiveness and efficiency of our approach.

[1]  R. R. Joshi,et al.  A new approach to map matching for in-vehicle navigation systems: the rotational variation metric , 2001, ITSC 2001. 2001 IEEE Intelligent Transportation Systems. Proceedings (Cat. No.01TH8585).

[2]  Xin Wang,et al.  Personalized travel route recommendation using collaborative filtering based on GPS trajectories , 2018, Int. J. Digit. Earth.

[3]  Xing Xie,et al.  Geography-Aware Sequential Location Recommendation , 2020, KDD.

[4]  Jingyuan Wang,et al.  Deep Trajectory Recovery with Fine-Grained Calibration using Kalman Filter , 2019, IEEE Transactions on Knowledge and Data Engineering.

[5]  John Krumm,et al.  Hidden Markov map matching through noise and sparseness , 2009, GIS.

[6]  Haifeng Wang,et al.  ConSTGAT: Contextual Spatial-Temporal Graph Attention Network for Travel Time Estimation at Baidu Maps , 2020, KDD.

[7]  Chung Choo Chung,et al.  Sequence-to-Sequence Prediction of Vehicle Trajectory via LSTM Encoder-Decoder Architecture , 2018, 2018 IEEE Intelligent Vehicles Symposium (IV).

[8]  Yike Guo,et al.  Deep Sequence Learning with Auxiliary Information for Traffic Prediction , 2018, KDD.

[9]  Fuzhen Zhuang,et al.  Modelling of Bi-Directional Spatio-Temporal Dependence and Users' Dynamic Preferences for Missing POI Check-In Identification , 2019, AAAI.

[10]  Zheng Wang,et al.  HetETA: Heterogeneous Information Network Embedding for Estimating Time of Arrival , 2020, KDD.

[11]  Zhifeng Bao,et al.  Online Anomalous Trajectory Detection with Deep Generative Sequence Modeling , 2020, 2020 IEEE 36th International Conference on Data Engineering (ICDE).

[12]  Lukasz Kaiser,et al.  Attention is All you Need , 2017, NIPS.

[13]  Chao Zhang,et al.  DeepMove: Predicting Human Mobility with Attentional Recurrent Networks , 2018, WWW.

[14]  Kai Zheng,et al.  Calibrating trajectory data for spatio-temporal similarity analysis , 2014, The VLDB Journal.

[15]  Quoc V. Le,et al.  Sequence to Sequence Learning with Neural Networks , 2014, NIPS.

[16]  Stefano Secci,et al.  Estimating human trajectories and hotspots through mobile phone data , 2014, Comput. Networks.

[17]  Diansheng Guo,et al.  AttnMove: History Enhanced Trajectory Recovery via Attentional Network , 2021, AAAI.

[18]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[19]  Rich Caruana,et al.  Multitask Learning , 1998, Encyclopedia of Machine Learning and Data Mining.

[20]  Chengyang Zhang,et al.  Map-matching for low-sampling-rate GPS trajectories , 2009, GIS.

[21]  Yoshua Bengio,et al.  Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation , 2014, EMNLP.

[22]  Changsheng Li,et al.  Characterizing Driving Styles with Deep Learning , 2016, ArXiv.

[23]  Haris N. Koutsopoulos,et al.  Route travel time estimation using low-frequency floating car data , 2013, 16th International IEEE Conference on Intelligent Transportation Systems (ITSC 2013).

[24]  Wei Cao,et al.  When Will You Arrive? Estimating Travel Time Based on Deep Neural Networks , 2018, AAAI.

[25]  Tomoharu Iwata,et al.  Travel route recommendation using geotags in photo sharing sites , 2010, CIKM.

[26]  Yanhua Li,et al.  ST-SiameseNet: Spatio-Temporal Siamese Networks for Human Mobility Signature Identification , 2020, KDD.

[27]  Yisong Yue,et al.  Generating Long-term Trajectories Using Deep Hierarchical Networks , 2016, NIPS.

[28]  Xing Xie,et al.  An Interactive-Voting Based Map Matching Algorithm , 2010, 2010 Eleventh International Conference on Mobile Data Management.

[29]  Yoshua Bengio,et al.  Neural Machine Translation by Jointly Learning to Align and Translate , 2014, ICLR.

[30]  Zoubin Ghahramani,et al.  A Theoretically Grounded Application of Dropout in Recurrent Neural Networks , 2015, NIPS.

[31]  Yu Zheng,et al.  GeoMAN: Multi-level Attention Networks for Geo-sensory Time Series Prediction , 2018, IJCAI.

[32]  Christian S. Jensen,et al.  Deep Representation Learning for Trajectory Similarity Computation , 2018, 2018 IEEE 34th International Conference on Data Engineering (ICDE).