The viewer is disabled because this dataset repo requires arbitrary Python code execution. Please consider
removing the loading script and relying on automated data support (you can use convert_to_parquet from the datasets library). If this is not possible, please open a discussion for direct help.
Dataset title: Porto Taxi.
Dataset authors: Julia Moska, Oleksii Furman, Kacper Kozaczko, Szymon Leszkiewicz, Jakub Polczyk, Piotr Gramacki, Piotr Szymański - Wrocław University of Science and Technology
Data authors: Generated from 442 taxis operating in Porto, Portugal. Data collected for the Prediction Challenge, ECML PKDD 2015, by Luis Moreira-Matias, Michel Ferreira, Joao Mendes-Moreira, L. L., and J. J.
Date range: From July 1, 2013, to June 30, 2014.
Research methodology and tools used for data acquisition: Originally collected to support improvements in electronic taxi dispatch systems. It captures highly detailed GPS trajectory data from 442 taxis. The datasets were preprocessed to remove incomplete entries and spatial outliers, such as records falling outside defined city boundaries. For benchmarking purposes, the data was standardized and aggregated into H3 hexagonal grids of varying resolutions using the SRAI library to establish unified training and testing splits.
Data format and structure: Each record corresponds to a completed taxi trip and contains metadata (origin call type, taxi and trip IDs, day type) alongside the GPS trajectory encoded as a sequence of geographic coordinates. The used subset contains nearly 400,000 individual trajectories.
Data sharing and licensing conditions (license): CC BY 4.0.
Related Publications:
- Julia Moska, Oleksii Furman, Kacper Kozaczko, Szymon Leszkiewicz, Jakub Polczyk, Piotr Gramacki, and Piotr Szymański. 2025. OBSR: Open Benchmark for Spatial Representations. In Proceedings of the 33rd ACM International Conference on Advances in Geographic Information Systems (SIGSPATIAL '25). Association for Computing Machinery, New York, NY, USA, 670–681. https://doi.org/10.1145/3748636.3762783
Funding: This research was funded in whole or in part by the National Science Centre, Poland grants: Preludium: 2024/53/N/ST6/03296 and Sonata: 2024/55/D/ST6/03082
- Downloads last month
- 134