The widespread adoption of sensor-enabled and mobile ubiquitous devices has caused an avalanche of big data that is mostly geospatially tagged. Most cloud-based big data processing systems are designed for general-purpose workloads, neglecting spatial-characteristics. However, interesting analytics often seek answers for proximity-alike queries. We fill this gap by providing custom geospatial service layer atop of Apache Spark. To be more specific, we leverage Spark to design a custom spatial-aware partitioning method to boost geospatial query performances. Our results show that our patches outperform state-of-the-art implementations by significant fractions.
In-memory Spatial-Aware Framework for Processing Proximity-Alike Queries in Big Spatial Data / Al Jawarneh, Isam Mashhour; Bellavista, Paolo; Corradi, Antonio; Foschini, Luca; Montanari, Rebecca; Zanotti, Andrea. - ELETTRONICO. - 2018:(2018), pp. 8514950.1-8514950.6. (Intervento presentato al convegno 23rd IEEE International Workshop on Computer Aided Modeling and Design of Communication Links and Networks, CAMAD 2018 tenutosi a esp nel 2018) [10.1109/CAMAD.2018.8514950].
In-memory Spatial-Aware Framework for Processing Proximity-Alike Queries in Big Spatial Data
Al Jawarneh, Isam Mashhour;Bellavista, Paolo;Corradi, Antonio;Foschini, Luca;Montanari, Rebecca;
2018
Abstract
The widespread adoption of sensor-enabled and mobile ubiquitous devices has caused an avalanche of big data that is mostly geospatially tagged. Most cloud-based big data processing systems are designed for general-purpose workloads, neglecting spatial-characteristics. However, interesting analytics often seek answers for proximity-alike queries. We fill this gap by providing custom geospatial service layer atop of Apache Spark. To be more specific, we leverage Spark to design a custom spatial-aware partitioning method to boost geospatial query performances. Our results show that our patches outperform state-of-the-art implementations by significant fractions.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.