Deriving Application Level Relationships by Analysing the Run Time Behaviour of Simple and Complex SQL Queries

TL;DR

This paper presents a novel method to analyze SQL query runtime behavior to infer application-level table relationships, using primary/foreign key properties and query data, implemented in Java with experimental validation.

Contribution

It introduces a new approach to predict database table relationships by analyzing simple and complex SQL queries' runtime behavior and query structure.

Findings

Successfully implemented in Java with experimental results

Effective in identifying application-level table relationships

Works with both simple and complex SQL queries

Abstract

This paper describes a unique approach to perform application behavioral analysis for identifying how tables might be related to each other. The analysis techniques are based on the properties of primary and foreign keys and also the data present in their respective columns. We have also implemented the idea using JAVA and presented experimental results in Demo Section.This paper introduces a unique approach to predict the possible application level relationships in databases with the help of the application relationship analysis of simple and complex SQL queries. Complex SQL queries are those which contain multiple constraints at different levels of the database. In the process of deriving relations, we first parse the SQL statements and then analyse the parsed information to extract related information.