# Feature Extraction and Representation of Urban Road Networks Based on Travel Routes

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Related Work

#### 2.1. Traffic Incidents on Urban Road Networks

#### 2.2. Network Representation Learning

## 3. Feature Learning Framework

#### 3.1. Model Architecture

#### 3.1.1. Character-Level Convolutional Neural Network

#### 3.1.2. Long Short-Term Memory Neural Network

#### 3.1.3. StreetNode2VEC Model

- Pretreatments: Data cleaning, map matching, and other pretreatments are executed in this step. The longitude and latitude sequences of GPS are converted into a route composed of road nodes, and the nodes are put through one-hot encoding as inputs of the next step.
- Embedding module: The Char-Cnn embedding model is used in this step to prepare the data required by the Bi-LSTM layer. After being vectorized, the nodes on the route are transferred to the next step as inputs.
- Bi-LSTM module: The Bi-LSTM layer and the Maxpool layer are used in this step to extract the advanced features of data, the fully connected layer is adopted to adjust the output size of data, and eventually the binary classification results of the route are output.

#### 3.2. Detailed Model Settings

#### 3.2.1. Construction of Synthetic Samples

#### 3.2.2. Parameters of the Embedding Module

#### 3.2.3. Parameters of the Bi-LSTM Module

#### 3.2.4. Parameters of Training Methods

#### 3.3. Model Training

#### 3.3.1. Model Parameter Calibration

#### 3.3.2. Training Parameter Calibration

## 4. Case Study

#### 4.1. Dataset Process

#### 4.2. Experimental Setup

#### 4.3. Results of Visualization

#### 4.4. Results of Road Nodes Similarity Analysis

#### 4.5. Results of Road Nodes Link Prediction

## 5. Conclusions and Future Directions

## Author Contributions

## Funding

## Conflicts of Interest

## References

- World Health Organization. Global status report on road safety 2018: Summary. Available online: https://apps.who.int/iris/handle/10665/277370 (accessed on 15 November 2020).
- Sayed, T.; Rodriguez, F. Accident Prediction Models for Urban Unsignalized Intersections in British Columbia. Transp. Res. Rec. J. Transp. Res. Board
**1999**, 1665, 93–99. [Google Scholar] [CrossRef] - Greibe, P.; Rodriguez, F. Accident prediction models for urban roads. Accid. Anal. Prev.
**2003**, 35, 273–285. [Google Scholar] [CrossRef] - Mountain, L.; Fawaz, B.; Jarrett, D. Accident prediction models for roads with minor junctions. Accid. Anal. Prev.
**1996**, 28, 695–707. [Google Scholar] [CrossRef] - Lee, J.; Abdel-Aty, M.; Jiang, X. Multivariate crash modeling for motor vehicle and non-motorized modes at the macroscopic level. Accid. Anal. Prev.
**2015**, 78, 146–154. [Google Scholar] [CrossRef] - Lee, J.; Abdel-Aty, M.; Cai, Q.; Wang, L.; Huang, H. Integrated modeling approach for non-motorized mode trips and fatal crashes in the framework of transportation safety planning. Transp. Res. Rec.
**2018**, 2672, 49–60. [Google Scholar] [CrossRef] - Mussone, L.; Ferrari, A.; Oneta, M. An analysis of urban collisions using an artificial intelligence model. Accid. Anal. Prev.
**1999**, 31, 705–718. [Google Scholar] [CrossRef] - Singh, G.; Pal, M.; Yadav, Y.; Singla, T. Deep neural network-based predictive modeling of road accidents. Neural Comput. Appl.
**2020**, 32, 1–10. [Google Scholar] [CrossRef] - Anderson, T.K. Kernel density estimation and K-means clustering to profile road accident hotspots. Accid. Anal. Prev.
**2009**, 41, 359–364. [Google Scholar] [CrossRef] - Theofilatos, A. Incorporating real-time traffic and weather data to explore road accident likelihood and severity in urban arterials. J. Saf. Res.
**2017**, 61, 9–21. [Google Scholar] [CrossRef] - Cameron, M.H.; Elvik, R. Nilsson’s Power Model connecting speed and road trauma: Applicability by road type and alternative models for urban roads. Accid. Anal. Prev.
**2010**, 42, 1908–1915. [Google Scholar] [CrossRef] - Congiu, T.; Sotgiu, G.; Castiglia, P.; Azara, A.; Piana, A.; Saderi, L.; Dettori, M. Built environment features and pedestrian accidents: An Italian retrospective study. Sustainability
**2019**, 11, 1064. [Google Scholar] [CrossRef][Green Version] - Perozzi, B.; Al-Rfou, R.; Skiena, S. Deepwalk: Online learning of social representations. In Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, New York, NY, USA, 24–27 August 2014; pp. 701–710. [Google Scholar]
- Mohan, D.; Bangdiwala, S.I.; Villaveces, A. Urban street structure and traffic safety. J. Saf. Res.
**2017**, 62, 63–71. [Google Scholar] [CrossRef] [PubMed] - Xie, K.; Wang, X.; Ozbay, K.; Yang, H. Crash frequency modeling for signalized intersections in a high-density urban road network. Anal. Methods Accid. Res.
**2014**, 2, 39–51. [Google Scholar] [CrossRef] - Zeng, Q.; Huang, H. Bayesian spatial joint modeling of traffic crashes on an urban road network. Accid. Anal. Prev.
**2014**, 67, 105–112. [Google Scholar] [CrossRef] [PubMed] - Demasi, F.; Loprencipe, G.; Moretti, L. Road safety analysis of urban roads: Case study of an Italian municipality. Safety
**2018**, 4, 58. [Google Scholar] [CrossRef][Green Version] - Lee, J.; Abdel-Aty, M.; Huang, H.; Cai, Q. Transportation safety planning approach for pedestrians: An integrated framework of modeling walking duration and pedestrian fatalities. Transp. Res. Rec.
**2019**, 2673, 898–906. [Google Scholar] [CrossRef] - Lee, J.; Abdel-Aty, M.; Choi, K.; Huang, H. Multi-level hot zone identification for pedestrian safety. Accid. Anal. Prev.
**2015**, 76, 64–73. [Google Scholar] [CrossRef] - Cantisani, G.; Moretti, L.; De Andrade Barbosa, Y. Safety problems in urban cycling mobility: A quantitative risk analysis at urban intersections. Safety
**2019**, 5, 6. [Google Scholar] [CrossRef][Green Version] - Cantisani, G.; Moretti, L.; De Andrade Barbosa, Y. Risk analysis and safer layout design solutions for bicycles in four-leg urban intersections. Safety
**2019**, 5, 24. [Google Scholar] [CrossRef][Green Version] - Lee, J. Development of Traffic Safety Zones and Integrating Macroscopic and Microscopic Safety Data Analytics for Novel Hot Zone Identification. Ph.D. Thesis, University of Central Florid, Orlando, FL, USA, 2014. [Google Scholar]
- Lee, J.; Abdel-Aty, M.; Jiang, X. Development of zone system for macro-level traffic safety analysis. J. Transp. Geogr.
**2014**, 38, 13–21. [Google Scholar] [CrossRef] - Lee, J.; Abdel-Aty, M.; Choi, K. Analysis of residence characteristics of at-fault drivers in traffic crashes. Saf. Sci.
**2014**, 68, 6–13. [Google Scholar] [CrossRef] - Xie, B.; An, Z.; Zheng, Y.; Li, Z. Incorporating transportation safety into land use planning: Pre-assessment of land use conversion effects on severe crashes in urban China. Appl. Geogr.
**2019**, 103, 1–11. [Google Scholar] [CrossRef] - Soltani, A.; Askari, S. Exploring spatial autocorrelation of traffic crashes based on severity. Inj. Int. J. Care Inj.
**2017**, 48, 637–647. [Google Scholar] [CrossRef] [PubMed] - Abdel-Aty, M.; Lee, J.; Siddiqui, C.; Choi, K. Geographical unit based analysis in the context of transportation safety planning. Transp. Res. Part A Policy Pract.
**2013**, 49, 62–75. [Google Scholar] [CrossRef] - Stipancic, J.; Miranda-Moreno, L.; Saunier, N.; Labbe, A. Network screening for large urban road networks: Using GPS data and surrogate measures to model crash frequency and severity. Accid. Anal. Prev.
**2019**, 125, 290–301. [Google Scholar] [CrossRef] [PubMed] - Cai, Q.; Abdel-Aty, M.; Sun, Y.; Lee, J.; Yuan, J. Applying a deep learning approach for transportation safety planning by using high-resolution transportation and land use data. Transp. Res. Part A Policy Pract.
**2019**, 127, 71–85. [Google Scholar] [CrossRef] - Marshall, W.E.; Garrick, N.W. Does street network design affect traffic safety? Accid. Anal. Prev.
**2011**, 43, 769–781. [Google Scholar] [CrossRef] - Bengio, Y.; Ducharme, R.; Vincent, P.; Jauvin, C. A neural probabilistic language model. J. Mach. Learn. Res.
**2003**, 3, 1137–1155. [Google Scholar] - Collobert Ronan, W.J. A unified architecture for natural language processing: Deep neural networks with multitask learning. In Proceedings of the 25th International Conference on Machine Learning, Helsinki, Finland, 5–9 July 2008; pp. 160–167. [Google Scholar] [CrossRef]
- Mikolov, T.; Sutskever, I.; Chen, K.; Corrado, G.S.; Dean, J. Distributed representations of words and phrases and their compositionality. In Advances in Neural Information Processing Systems; MIT Press: Cambridge, MA, USA, 2013; pp. 3111–3119. [Google Scholar]
- Jian, T.; Meng, Q.; Mingzhe, W.; Ming, Z.; Jun, Y.; Qiaozhu, M. LINE: Large-scale Information Network Embedding. In Proceedings of the 24th International Conference on World Wide Web, Florence, Italy, 18–22 May 2015; pp. 1067–1077. [Google Scholar] [CrossRef][Green Version]
- Grover Aditya, L.J. node2vec: Scalable feature learning for networks. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA, 24–27 August 2016; pp. 855–864. [Google Scholar] [CrossRef][Green Version]
- Perozzi, B.; Kulkarni, V.; Skiena, S. Walklets: Multiscale graph embeddings for interpretable network classification. arXiv
**2016**, arXiv:1605.02115. [Google Scholar] - Giang Hoang, N.; John Boaz, L.; Ryan, A.R.; Nesreen, K.A.; Eunyee, K.; Sungchul, K. Continuous-Time Dynamic Network Embeddings. In Proceedings of the Companion Proceedings of the Web Conference 2018, Lyon, France, 23–27 April 2018; pp. 969–976. [Google Scholar] [CrossRef][Green Version]
- Zhang, X.; Zhao, J.; LeCun, Y. Character-level convolutional networks for text classification. In Advances in Neural Information Processing Systems; MIT Press: Cambridge, MA, USA, 2015; pp. 649–657. [Google Scholar]
- Sun, X.; Guo, J.; Ding, X.; Liu, T. A general framework for content-enhanced network representation learning. arXiv
**2016**, arXiv:1610.02906. [Google Scholar] - Graves, A.; Schmidhuber, J. Framewise phoneme classification with bidirectional LSTM and other neural network architectures. Neural Netw.
**2015**, 18, 602–610. [Google Scholar] [CrossRef] [PubMed] - Mikolov, T.; Chen, K.; Corrado, G.; Dean, J. Efficient estimation of word representations in vector space. arXiv
**2013**, arXiv:1301.3781. [Google Scholar] - Goodfellow, I.; Pouget-Abadie, J.; Mirza, M.; Xu, B.; Warde-Farley, D.; Ozair, S.; Courville, A.; Bengio, Y. Generative adversarial nets. In Advances in Neural Information Processing Systems; MIT Press: Cambridge, MA, USA, 2014; pp. 2672–2680. [Google Scholar]
- Pennington, J.; Socher, R.; Manning, C.D. GloVe: Global Vectors for Word Representation. In Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing, Doha, Qatar, 26–28 October 2014; pp. 1532–1543. [Google Scholar]
- Lutz, P. Early stopping-but when? In Neural Networks: Tricks of the Trade; Springer: Berlin/Heidelberg, Germany, 1998; pp. 55–69. [Google Scholar]
- Kaggle. ECML/PKDD 15: Taxi Trajectory Prediction (I). Available online: https://www.kaggle.com/c/pkdd-15-predict-taxi-service-trajectory-i/ (accessed on 15 November 2020).
- Maaten, L.V.D.; Hinton, G. Visualizing data using t-SNE. J. Mach. Learn. Res.
**2008**, 9, 2579–2605. [Google Scholar] - Kipf, T.; Welling, M. Semi-Supervised Classification with Graph Convolutional Networks. arXiv
**2016**, arXiv:1609.02907. [Google Scholar] - O’Flaherty, C.A. Highways. In Traffic Planning and Engineering; Edward Arnold: London, UK, 1986; Volume 1. [Google Scholar]

**Figure 3.**Calibration of training parameters. (

**a**) Calibration of batch size. (

**b**) Calibration of learning rate.

**Figure 5.**Road nodes visualization of Porto road network by StreetNode2VEC. (

**a**) 2D embedding of the road nodes. (

**b**) Road node visualization with label colors.

**Figure 6.**ROC curve of models on operators. (

**a**) ROC curve of models on operator Average. (

**b**) ROC curve of models on operator Hadamard. (

**c**) ROC curve of models on operator L1. (

**d**) ROC curve of models on operator L2.

Element-Wise Product | Formula of Element-Wise Product |
---|---|

Average | $(u+v)/2$ |

Hadamard | $u\ast v$ |

L1 | $|u+v|$ |

L2 | ${(u-v)}^{2}$ |

$cos(A,B)=0.919$ $cos(A,C)=-0.460$ $cos(A,D)=0.004$ | |

$cos(A,B)=0.997$ $cos(A,C)=-0.933$ $cos(A,D)=0.001$ |

Model Name | AUC Score | |||
---|---|---|---|---|

Average | Hadamard | L1 | L2 | |

Node2vec | 0.503 | 0.749 | 0.711 | 0.656 |

GCN | 0.609 | 0.825 | 0.752 | 0.711 |

StreetNode2VEC | 0.772 | 0.785 | 0.818 | 0.803 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Huang, S.; Shao, C.; Li, J.; Yang, X.; Zhang, X.; Qian, J.; Wang, S. Feature Extraction and Representation of Urban Road Networks Based on Travel Routes. *Sustainability* **2020**, *12*, 9621.
https://doi.org/10.3390/su12229621

**AMA Style**

Huang S, Shao C, Li J, Yang X, Zhang X, Qian J, Wang S. Feature Extraction and Representation of Urban Road Networks Based on Travel Routes. *Sustainability*. 2020; 12(22):9621.
https://doi.org/10.3390/su12229621

**Chicago/Turabian Style**

Huang, Shichen, Chunfu Shao, Juan Li, Xiong Yang, Xiaoyu Zhang, Jianpei Qian, and Shengyou Wang. 2020. "Feature Extraction and Representation of Urban Road Networks Based on Travel Routes" *Sustainability* 12, no. 22: 9621.
https://doi.org/10.3390/su12229621