Because of the large amount trajectory data produced by mobile devices, there is an increasing need for mechanisms to extract knowledge from this data. Most existing works have focused on geometric properties trajectories, but recently emerged concept semantic in which background geographic information integrated sample points. In new concept, trajectories are observed as a set stops and moves, where most important parts trajectory. Stops moves been computed testing intersections with...

The collection of moving object data is becoming more and common, therefore there an increasing need for the efficient analysis knowledge extraction these in different application domains. Trajectory are normally available as sample points, do not carry semantic information, which fundamental importance comprehension data. Therefore, trajectory becomes expensive from a computational point view complex user's perspective. Enriching trajectories with geographical information may simplify...

In this paper, we present a unified approach for abnormal behavior detection and group analysis in video scenes. Existing approaches do either use trajectory-based or pixel-based methods. Unlike these approaches, propose an integrated pipeline that incorporates the output of object trajectory inference. This enables to detect behaviors related speed direction trajectories, as well complex finer motion each object. By applying our on three different data sets, show is able several types with...

Abstract Several works have been proposed in the last few years for raw trajectory data analysis, and some attempts made to define trajectories from a more semantic point of view. Semantic analysis has received significant attention recently, but formal definition trajectory, set aspects that should be considered semantically enrich conceptual model integrating these broad sense is still missing. This article presents named CONSTAnT , which defines most important trajectories. We believe...

Existing works for semantic trajectory data analysis have focused on the intersection of trajectories with application important geographic information and use speed to find interesting places. In this paper we present a novel approach places in trajectories, considering variation direction as main aspect. The proposed has been validated real associated oceanic fishing vessels, objective automatically where vessels develop activities. Results demonstrated that method is very appropriate...

Abstract Most existing approaches aiming at measuring trajectory similarity are focused on two‐dimensional sequences of points, called raw trajectories. However, recent proposals have used background geographic information and social media data to enrich these trajectories with a semantic dimension, giving rise the concept Only few works proposed measures for or multidimensional sequences, having limitations such as predefined weight dimensions, sensitivity noise, tolerance gaps different...

Abstract For many years trajectory data have been treated as sequences of space‐time points or stops and moves. However, with the explosion Internet Things flood big generated on Internet, such weather channels social network interactions, which can be used to enrich mobility data, trajectories become more complex, multiple heterogeneous dimensions. The main challenge is how integrate all this information trajectories. In article we introduce a new concept trajectory, called aspect propose...

Mobile devices are becoming very popular in recent years, and large amounts of trajectory data generated by these devices. Trajectories left behind cars, humans, birds or other objects a new kind which can be useful the decision making process several application domains. These data, however, normally available as sample points, therefore have little no semantics. The analysis knowledge extraction from points is difficult user's point view, there an emerging need for models, manipulation...

Abstract The large amount of semantically rich mobility data becoming available in the era big has led to a need for new trajectory similarity measures. In context multiple‐aspect trajectories, where are enriched with several semantic dimensions, current state‐of‐the‐art approaches present some limitations concerning relationships between attributes and their semantics. Existing works either too strict, requiring match on all attributes, or flexible, considering as independent. this article...

Abstract Enormous quantities of trajectory data are collected from many sources, such as GPS devices and mobile phones, sequences spatio‐temporal points. These can be used in application domains traffic management, urban planning, tourism, bird migration, so on. Raw data, generated by have very little or no semantics, most applications a higher level abstraction is needed to exploit these for decision making. Although several different methods been proposed far querying mining, there...

Trajectory data analysis and mining require distance similarity measures, the quality of their results is directly related to those measures. Several measures originally proposed for time-series were adapted work with trajectory data, but these approaches developed well-behaved that usually do not have uncertainty heterogeneity introduced by sampling process obtain trajectories. More recently, specifically they rely on simplistic movement representations, such as linear interpolation. In...

Several methods for trajectory classification build models exploring global features, such as the average and standard deviation of speed acceleration, but some applications these features may not be best to determine class. Other works explore local applying partition discretization, that lose important movement information could discriminate In this work we propose a new method, called Movelets, discover relevant subtrajectories without need predefined criteria either or discretization. We...

For many years trajectory similarity research has focused on raw trajectories, considering only space and time information. With the semantic enrichment, emerged need for measures that support space, time, semantics. Although some deal with all these dimensions, they consider stops, ignoring moves. We claim that, applications, movement between stops is as important must be considered in analysis. In this article, we propose SMSM, a novel measure trajectories considers both evaluate SMSM...

Abstract During the last few years volumes of data that synthesize trajectories have expanded to unparalleled quantities. This growth is challenging traditional trajectory analysis approaches and solutions are sought in other domains. In this work, we focus on compression techniques with intention minimize size data, while, at same time, minimizing impact methods. To extent, evaluate five lossy algorithms: Douglas-Peucker (DP), Time Ratio (TR), Speed Based (SP), (TR_SP) (SP_TR). The...

Abstract Research on trajectory behavior has increased significantly in the last few years. The focus been search for patterns considering movement of moving object space and time, essentially looking similar geometric properties dense regions. This paper proposes an algorithm to detect a new kind pattern that identifies when is avoiding specific spatial regions, such as security cameras. called avoidance. was evaluated with real data achieved very good results.

In frequent geographic pattern mining a large amount of patterns is well known priori. This paper presents novel approach for without associations that are previously as non- interesting. Geographic dependences eliminated during the set generation using prior knowledge. After dependence elimination maximal generalized sets computed to remove redundant sets. Experimental results show significant reduction both number and computational time patterns.

Abstract Many association rule‐mining algorithms have been proposed in the last few years. Their main drawback is huge amount of generated patterns. In spatial rule mining, besides large rules, many are well‐known geographic domain associations explicitly represented database schemas. Existing only considered data, while schema has not considered. The result that also schemas extracted by algorithms. With aim to reduce number patterns and this paper presents a summary results novel approach...

The large amount of patterns generated by frequent pattern mining algorithms has been extensively addressed in the last few years. In geographic mining, besides patterns, many are well known domain associations. Existing do not warrant elimination all dependences since no prior knowledge is used for this purpose. This paper presents a two step method without associations that previously as non-interesting. first input space reduced much possible. far we know still most efficient to reduce...