Navigating vast datasets can be a computational nightmare. This paper introduces a novel solution: the multidimensional binary search tree (k-d tree), a data structure designed for efficient storage and retrieval of information through associative searches. The paper defines the structure of k-d trees and provides illustrative examples. This paper showcases the efficiency of k-d trees in terms of storage requirements. A single k-d tree can efficiently handle various types of queries. Utility algorithms for insertion, deletion, and optimization are developed. The paper proves the average running times for these algorithms in an n-record file, which is critical for assessing their practical applicability. With demonstrated running times that surpass existing algorithms, k-d trees hold significant promise for diverse applications. Examples of potential uses are provided, highlighting the practical relevance of this theoretical work. While the paper's focus is primarily theoretical, it paves the way for future research and implementation of k-d trees in real-world scenarios, particularly for managing and querying large, multidimensional datasets.
Published in Communications of the ACM, a leading journal in computer science, this paper's focus on data structures and algorithms is highly relevant to the journal's scope. Given the Communications of the ACM's focus on practical applications of computer science research, this paper's discussion of potential uses for k-d trees aligns well with the journal's emphasis on bridging theory and practice.