GCT-TTE

Welcome to the official repo of the GCT-TTE model -- transformer-based travel time estimation algorithm. Here we present the source code of the pipeline and demo application.

You can access the inference of our model at gctte.online

arXiv PDF: https://arxiv.org/abs/2306.04324

Prerequisites

Backend: please use application/requirements.txt in order to compile the environment for the application.

Model: the experiments were conducted with CUDA 10.1 and torch 1.8.1. The following libraries must be compatible with this software setup:

- torch-cluster==1.6.0
- torch-geometric==2.1.0.post1
- torch-scatter==2.0.8
- torch-sparse==0.6.12
- torch-spline-conv==1.2.1

All other external libraries, which do not depend on torch and CUDA versions, are mentioned in /model/requirements.txt.

Local tests

Launch instructions are provided in the README file of the /model directory.

Datasets

We provide two datasets corresponding to the cities of Abakan and Omsk. For each of these datasets, there are two types of target values -- real travel time (considered in this study) and real length of trip.

Road network	Trips
Abakan Omsk Nodes 65524 231688 Edges 340012 1149492 Clustering 0.5278 0.53 Usage median 12 8	Abakan Omsk Trips number 121557 767343 Coverage 53.3% 49.5% Average time 427 sec 608 sec Average length 3604 m 4216 m

Provided data could be used for research purposes only. If you want to incorporate the graph data in your study, please send a request to [email protected]. The image extension can be accesed via https://sc.link/Mw9kP (Abakan) and https://sc.link/5QWBq (Omsk).

License

Established code released as open-source software under the MIT license.

Contact us

If you have some questions about the code, you are welcome to open an issue, I will respond to that as soon as possible.

Citation

@Article{Mashurov2024,
        author={Mashurov, Vladimir and Chopuryan, Vaagn and Porvatov, Vadim and Ivanov, Arseny and Semenova, Natalia},
        title={GCT-TTE: graph convolutional transformer for travel time estimation},
        journal={Journal of Big Data},
        year={2024},
        month={Jan},
        day={13},
        volume={11},
        number={1},
        pages={15},
        doi={10.1186/s40537-023-00841-1},
        url={https://doi.org/10.1186/s40537-023-00841-1}
}