A Pictorial Performance Comparison of Spatial Indexes

Data is ultimately stored on storage media, where indexes are used to store and retrieve objects. The needs for optimized storage and fast retrieval of data have gained because of big data, spatial (multi-dimensional). Tree-based indexes are usually compared by conducting numerical comparisons for quality factors such as the overlap percentage and the number of disk accesses needed to retrieve a certain amount of data. The graphical representation of actual trees produced by an index expresses the distribution of objects at each tree level and uncovers the anomalies in nodes coverage in addition to the overlap areas between nodes. This article uses the graphical representation of index trees concerning two datasets to give a clear understanding of each index performance by revealing how index-nodes are distributed, shaped, and where overlap areas are concentrated.