Using Text Classification Techniques to identify anti-patterns in SQL queries

Abstract

A significant issue with utilizing social information bases, is formulating and writing proficient SQL
queries. Some basic blunders known as anti-patterns are quite common in SQL queries and can genuinely affect its
execution time and at times, the database's general execution. This paper manifests machine learning methods and
identifies anti-patterns by approaching the issue as a text classification issue. Our outcome is a model dependent on a
convolutional neural network that can be utilized to order a SQL query into zero, one or numerous anti-pattern classes.