Recent advances on tracking technologies enable the collection of spatio-temporal data in the form of trajectories. The analysis of such data can convey knowledge in prominent applications, and mining groups of moving objects turns out to be a valuable mean to model their movement. Existing approaches pay particular attention in groups where objects are close and move together or follow similar trajectories by assuming that movement cannot change over time. Instead, we observe that groups can be of interest also when objects are spatially distant and have different but inter-related movements: objects can start from different places and join together to move towards a common location. To take into account inter-related movements, we have to analyze the objects jointly, follow their respective movements and consider changes of movements over time. Motivated by this, we introduce the notion of communities and propose a computational solution to discover them. The method is structured in three steps. The first step performs a feature extraction technique to elicit the inter-related movements between the objects. The second one leverages a tree-structure in order to group objects with similar inter-related movements. In the third step, these groupings are used to mine communities as groups of objects which exhibit inter-related movements over time. We evaluate our approach on real data-sets and compare it with existing algorithms.

Mining Trajectory Data for Discovering Communities of Moving Objects

LOGLISCI, CORRADO;MALERBA, Donato;
2014

Abstract

Recent advances on tracking technologies enable the collection of spatio-temporal data in the form of trajectories. The analysis of such data can convey knowledge in prominent applications, and mining groups of moving objects turns out to be a valuable mean to model their movement. Existing approaches pay particular attention in groups where objects are close and move together or follow similar trajectories by assuming that movement cannot change over time. Instead, we observe that groups can be of interest also when objects are spatially distant and have different but inter-related movements: objects can start from different places and join together to move towards a common location. To take into account inter-related movements, we have to analyze the objects jointly, follow their respective movements and consider changes of movements over time. Motivated by this, we introduce the notion of communities and propose a computational solution to discover them. The method is structured in three steps. The first step performs a feature extraction technique to elicit the inter-related movements between the objects. The second one leverages a tree-structure in order to group objects with similar inter-related movements. In the third step, these groupings are used to mine communities as groups of objects which exhibit inter-related movements over time. We evaluate our approach on real data-sets and compare it with existing algorithms.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11586/65659
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 6
  • ???jsp.display-item.citation.isi??? ND
social impact