METHOD AND DEVICE FOR EVALUATING METRIC SPACE PARTITIONING MODE, COMPUTER DEVICE AND STORAGE MEDIUM

摘要

A method and device (200) for evaluating a metric space partitioning mode, a computer device and a storage medium, the method comprising: performing metric space partitioning on data in a database to obtain n normal vectors (S101); inputting n normal vectors for partitioning a hyperplane, and a range search radius, r (S102); using a point selection algorithm to select n pivots (S103); using the distance between the data and the pivots as coordinates, and mapping the data in the metric space into two-dimensional vector data (S104); performing complete partitioning on the data and calculating the data that fall into an r-neighborhood (S105); outputting the number of data that fall into the r-neighborhood and outputting an evaluation result of the partitioning mode according to the number (S106).The method does not require a range search for all the data in the database, and only requires calculation, when an index is created, of the number of data that fall into the vicinity of the partitioning boundary and cannot be excluded. The time cost of evaluation is greatly reduced, and the efficiency is improved.