Enhancement for graph operations in relational database for criminal intelligence domain. (24th November 2020)

• Record Type: Journal Article
• Authors: Piech, Mateusz; Łoś, Marcin; Marcjan, Robert
• Is Part Of: International journal of intelligent information and database systems. Volume 14:Number 1(2021)
• Page Start: 49
• Page End: 66
• Publication Date: 2020-11-24
• Subjects: criminal data -- common table expression -- CTE -- graph database -- JSON -- PostgreSQL; Database management -- Computer programs -- Periodicals; Information retrieval -- Computer programs -- Periodicals; Information storage and retrieval systems -- Computer programs -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Intelligent agents (Computer software) -- Periodicals; 006.33
• Languages: English
• ISSNs: 1751-5858

Abstract

The aim of the research was to improve the performance of graph operations in relational database for semi-structured data. First of all, this required us to select a model that allows storing semi-structured data with relations. For this purpose, we selected one of the existing solutions, which stores semi-structured data in a model built using JSON as a native type and adapted it to our requirements. Secondly, we created an algorithm based on common table expression with recursion with the extension of JSON to facilitate more complex graph analysis of data. We compared proposed solutions with Neo4j, chosen as a representative of graph databases. The results show that we obtained an improvement in execution performance in some cases. Although we focused our use cases on criminal intelligence domain, the research output can be applied to every domain with semi-structured data.