Indian Journal of Science and Technology

Abstract

Objectives: In this paper, we show the importance of considering the order of occurrence of the events of a trajectory (chronology) as a relevant factor to calculate the similarity among trajectories and propose some methods. Methods/ Analysis: We proposed four methods with different degrees of strictness depending on the contiguity of the visited sites, which can be applied according to the interests of the analyst and the application domain. We made experiments from real data to evaluate the behavior of each method and thus show the advantages and disadvantages of each one. Findings: The similarity measure proposed gives a flexible way to calculate chronological similarity between users. This could be applied in fields such as marketing, pattern analysis, and clustering. Novelty/Improvements: Our methods consider different chronological aspects for finding the similarity of trajectories. It can be used with current methods to consider visited sites and activities performed in these sites.

Keywords: Semantic Trajectories, Similarity Measures, Trajectory Similarity, Trajectory Chronology