Scalable Spatial Data Science for Social Scientists
The proliferation of spatial data from modern and increasingly prevalent technologies has resulted in large datasets, ripe for extracting insightful knowledge that can drive many applications.
Overview
Abstract
The proliferation of spatial data from modern and increasingly prevalent technologies has resulted in large datasets, ripe for extracting insightful knowledge that can drive many applications. The most common datasets indexed by Google Dataset Search come from geosciences and social sciences, with 14 million datasets representing 45.2% of all indexed datasets, accessed by one-third of the users, and dominated by spatial data. This, in turn, necessitates the deployment of scalable and expressive analysis tools capable of extracting significant global insights. In this landscape, spatial statistical analysis emerges as a critical tool enabling social scientists to analyze large spatial data, while maintaining statistical significance in their results. However, an evident gap remains, as the current roster of spatial data science tools either support spatial statistical analysis but falter when handling large-scale datasets, or can manage big data but struggle to articulate spatial statistical queries effectively. Our research bridges the two worlds, exploring innovative query processing techniques that amplify both query expressiveness and scalability. Moreover, our work forms part of an expansive vision for identifying common utilities for spatial statistical analysis that merit system-level support, foreseeing a more profound integration between big data systems and spatial data science tools.t.
Brief Biography
Amr Magdy is a Professor of Computer Science and Engineering and a co-founding faculty member of the Center for Geospatial Sciences at UC Riverside. His research interests include big data management, spatial data management, large-scale data analytics, indexing, and main-memory management. His research has been published in prestigious research venues, including ACM SIGSPATIAL, VLDB, VLDB Journal, ACM SIGMOD, ACM TSAS, IEEE ICDE, and IEEE TKDE. Amr's research is recognized as the best paper runner-up in IEEE MDM 2023, among the best papers of SSTD 2023, ACM SIGSPATIAL 2019, and IEEE ICDE 2014, and has been incubated by several industrial collaborators in the Middle East and USA. Amr has received several research awards, including the Google-CAHSI and NSF CAREER awards in 2023.