Generating spatio-temporal streaming trajectories

Mešković, Emir; Osmanović, Dario; Galić, Zdravko; Baranović, Mirta

doi:10.1109/mipro.2014.6859738

Cited by 1 publication

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…TelegraphCQ (online) [14] Possible link for offline with Hermes(PostGis) [74] -Sharing paradigm (Psoup) [16] -Adaptive processing (eddy) [10] -Dynamic scheduling [86] -ST Sampling [76] -Windows aggregation [73] -Windows at different granularity level [72] -Index at different granularity level [77] Flux [81] SCUBA [62] CAPE [78] -Plan migration [97] -Adaptive scheduling [85] -Cluster sharing paradigm [62] -Clustersheddy [63] -Aggregation D-CAPE [84] GeoInsight [48] Microsoft StreamInsight [7] -Fusing horizontal & vertical [7] -Stream partitionning knn range queries [56] -Event-based [7] -Views derived from archive in-memory [48] -Native support for ST stream [6] Infosphere Streams ITS [12] SPADE [7] -scheduling component [89] -operator fusing [49] -basic PE (Processing Element) -map-matching [6] -shortest path -Datafaflow [12] Zaghreb laboratory works -TelegraphCQ -Implementation in java -General framework for MO [30] -Uncertainty handling -Trajectory buffering [55] thors proposes to deal with moving objects in a online way only by processing data in-memory, but this work doesn't take care of distribution aspects.…”

Section: Kostas Patroumpas Workmentioning

confidence: 99%

Design principles of a stream-based framework for mobility analysis

Salmon

Ray

2016

Geoinformatica

View full text Add to dashboard Cite

Trajectory analysis is of crucial importance in several fields as social analysis, zoology, climatology or traffic monitoring. Over the last decade, the number of mobile systems and devices recording their positions has grown significantly generating a deluge of spatial and temporal data to analyze. This increasing volume of data raises numerous issues in terms of storage, processing and extraction of information. Previous works considering movement analysis have been mainly oriented towards either archived data processing and mining or continuous handling of incoming streams. The research developed in this paper introduces the design principles of a holistic approach combining real-time processing and archived data analysis to process mobility data "on the fly". This solution aims to provide better results comparing to both purely offline and online approaches. This research considers distributed data and processing to be more efficient. The design principles are applied to maritime traffic analysis and a few representative examples are introduced to demonstrate the relevance of our approach.

show abstract