Which Way to Go - Finding Frequent Trajectories Through Clustering

• Type: Article in International Conference Proceedings Book
• Date: 2023

Authors

Thiago Andrade Silva (Author) Other; João Gama (Author) FEP

Conference proceedings International

Title: Discovery Science (DS)

Pages: 460-473

26th International Conference on Discovery Science, DS 2023

Porto, 9 October 2023 through 11 October 2023

Indexing

Scopus - 4 Citations

Other information

Authenticus ID: P-00Z-540

DOI: 10.1007/978-3-031-45275-8_31

Abstract (EN): Trajectory clustering is one of the most important issues in mobility patterns data mining. It is applied in several cases such as hot-spots detection, urban transportation control, animal migration movements, and tourist visiting routes among others. In this paper, we describe how to identify the most frequent trajectories from raw GPS data. By making use of the Ramer-Douglas-Peucker (RDP) mechanism we simplify the trajectories in order to obtain fewer points to check without losing information. We construct a similarity matrix by using the Fréchet distance metric and then employ density-based clustering to find the most similar trajectories. We perform experiments over three real-world datasets collected in the city of Porto, Portugal, and in Beijing China, and check the results of the most frequent trajectories for the top-k origins x destinations for the moves. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Language: English

Type (Professor's evaluation): Scientific

No. of pages: 13