Title: Investment analytics using association rule mining (Finassociations)

Authors: Elif Kartal; M. Erdal Balaban; Zeki Özen

Addresses: Department of Management Information Systems, Faculty of Economics, İstanbul University, Türkiye ' Department of Business Administration (Emeritus), School of Business, İstanbul University, Türkiye ' Department of Management Information Systems, Faculty of Economics, İstanbul University, Türkiye

Abstract: This study aims to discover financial associations (relations) in (foreign) exchange rates, cryptocurrencies, and stocks using association rule mining (ARM). It demonstrates the applicability and success of ARM on alternative investment instruments over desired periods. A dynamic web application called 'Finassociations' was developed in this scope, allowing investors to use and discover ARM. They can use the desired filters to make investment decisions by generating rules for which investment instruments rise or fall together. The application dynamically retrieves current data from Yahoo Finance. This study is a dynamic and expanded update on the existing ones. The exemplary analyses utilised data spanning various periods, up to two years preceding October 9, 2022. According to the study results, significant and strong financial associations in three different investment groups can be obtained. Also, the results show that short-term financial data can be preferred over long-term financial data when examining associations between investment instruments.

Keywords: association rule mining; ARM; association rules; data mining; apriori; investment decisions; financial associations; finance; foreign exchange rates; cryptocurrencies; stocks.

DOI: 10.1504/IJCEE.2026.150997

International Journal of Computational Economics and Econometrics, 2026 Vol.16 No.1/2, pp.3 - 22

Received: 12 Dec 2023
Accepted: 15 Apr 2024
Published online: 08 Jan 2026 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article