System and method for classification of low relevance records in a database using instance-based classifiers and machine learning

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computing device to cause the computing device to:

• receive a request to delete a selected record in a database;
• in response to the receiving the request to delete the selected record, generate a representation of the selected record, delete the selected record in the database, and store the representation of the deleted selected record;
• in response to the storing the representation of the deleted selected record, determine a cluster from which the representation of the deleted selected record has a shortest determined distance, among a plurality of clusters, and assign the representation of the deleted selected record to the determined cluster;
• determine a distance between a record in the database and a nearest cluster among the plurality of clusters;
• and
• in response to the record being within a predetermined distance of the nearest cluster, determine that the record is a deletion candidate record.

A system, comprising:

• a hardware processor, a computer readable memory, and a computer readable storage medium associated with a computer device;
• program instructions of a database record deleter configured to, in response to a request to delete a selected record in a database, generate a vector representation of the selected record, delete the selected record in the database, and store the vector representation of the deleted selected record;
• program instructions of a cluster determiner configured to, in response to the database record deleter storing the vector representation of the deleted selected record, determine a cluster from which the vector representation of the deleted selected record has a shortest determined distance, among a plurality of clusters into which a plurality of vector representations of deleted records is partitioned;
• and
• program instructions of a deletion candidate determiner configured to determine a distance between a record in the database and a nearest cluster among the plurality of clusters into which the plurality of vector representations of deleted records is partitioned and, in response to the record being within a predetermined distance of the nearest cluster, determine that the record is a deletion candidate record,
• wherein the program instructions are stored on the computer readable storage medium for execution by the hardware processor via the computer readable memory.