Download presentation
Presentation is loading. Please wait.
Published byTyyne Niemelä Modified over 6 years ago
1
CLUSTER BY: A NEW SQL EXTENSION FOR SPATIAL DATA AGGREGATION
Chengyang Zhang and Yan Huang [chengyang, Department of Computer Science & Engineering University of North Texas Our Contributions Enables queries over spatial clusters in spatial databases Clusters are elevated into spatial objects on which spatial functions and predicates can be directly applied Challenges Efficient aggregation/elevation algorithms Query processing and optimization Extension to spatial streaming data v.s. Group By Cluster By Only classifies tuples with the same values Allow spatial points to be clustered based on spatial proximity Future Work: implementations Spatial functions do not accept discrete points as operands Clusters are elevated to spatial objects as the input to spatial functions Motivating Scenario Spatial Data Selection Clustering Polygonization Identify low income neighborhoods Clusters of homes with low income. House information is stored in a spatial table House SELECT Area(Polygon(location)) FROM House WHERE household_income < 30K CLUSTER BY location SELECT Area(Polygon(location)) FROM House WHERE household_income < 30K CLUSTER BY location SELECT AREA(Polygon(location)) FROM House WHERE household_income < 30K CLUSTER BY location
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.