Statistical database query using random sampling of records

摘要

A system and method for expediting database queries by using random sampling. Data associated with a database attribute is partitioned into multiple data classes using a query language grouping command. Each of the data classes is randomly sampled on an individual basis to obtain a corresponding number of class data samples, each of which are stored in a separate sample database table. Database queries are then applied to each of the class data samples in the sample database table.