Fuzzy association rule mining using binary particle swarm optimization: Application to cyber fraud analytics

In this paper, we developed a Binary Particle Swarm Optimization (BPSO) based fuzzy association rule miner to generate fuzzy association rules from a transactional database by formulating a combinatorial global optimization problem, without pre-defining minimum support and confidence unlike other conventional association miners. Goodness of fuzzy association rules is measured by a fitness function viz., the product of support and confidence. So as to demonstrate the effectiveness of our method, we implemented it to phishing detection domain. Based on the goodness of the rules obtained, we infer that our proposed algorithm can be used as a sound alternative to the fuzzy apriori algorithm.