A novel concurrent relational association rule mining approach

摘要: Abstract Data mining techniques are intensively used to uncover relevant patterns in large volumes of complex data which continuously extended with newly arrived instances. Relational association rules (RARs), a analysis and concept, have been introduced as an extension classical (ARs) for capturing various relationships between the attributes characterizing data. Due its NP-completeness, problem all interesting RARs within set is computationally difficult. As dimensionality be mined increases, algorithm Discovery Association Rules (DRAR) fails providing reasonable time. This paper introduces new approach named CRAR (Concurrent Rule mining) uses concurrency discovery process thus significantly reduces The effectiveness empirically validated on nine open source sets. reduction time when using against DRAR emphasizes that it can successfully applied practical scenarios.