A Comprehensive Review on Vertical and Horizontal Association Rule Mining Algorithms

Abstract

Association rule mining (ARM) is a fundamental technique in data mining, widely employed to uncover significant relationships and latent patterns within large‐scale transactional datasets. Since the introduction of the Apriori algorithm, ARM methodologies have evolved substantially to address the challenges presented by increasingly voluminous and complicated data, as well as emerging application requirements. Numerous algorithms, including ECLAT, FP‐Growth, tertius, utility‐based mining, sequential pattern mining, and graph mining approaches, have been designed to improve computing performance, scalability, and the quality of extracted patterns. In recent years, the advent of Big Data, distributed computing, explainable artificial intelligence (XAI), knowledge graphs, federated learning, and large language models (LLMs) has significantly expanded the scope of ARM beyond classic market basket analysis. This review thoroughly examines the progression of ARM, providing a comparative study of both classical and contemporary algorithms in terms of their search strategies, data representations, candidate generation approaches, computational characteristics, advantages, limitations, and application domains. The performance of ARM algorithms is assessed through a case study. Recent advancements in distributed and streaming ARM, graph pattern mining, high‐utility and sequential pattern mining, knowledge graph rule learning, and explainable rule discovery are also discussed. Emerging trends such as privacy‐preserving mining, AI‐assisted rule generation, and neuro‐symbolic knowledge discovery are also discussed to identify future research directions. Through integrating fundamental techniques with recent developments in artificial intelligence and large‐scale data analytics, this survey aims to guide researchers and practitioners in choosing suitable ARM algorithms and identifying viable avenues for future work within scalable, interpretable, and intelligent knowledge discovery.