International Journal of Engineering and Technology

Contemporary research in mining high utility itemsets from the databases faces two major challenges: exponential search space and database-dependent minimum utility threshold. The search space is very huge when number of distinct items and size of the database is very large. Data analysts must specify suitable minimum utility thresholds for their mining tasks though they may have no knowledge pertaining to their databases. To evade these problems, two approaches are presented to mine high utility itemsets from transaction databases with or without specifying minimum utility threshold by using genetic algorithm. To the best of our knowledge, this is the first work on mining high utility itemsets from transaction databases using Genetic Algorithm (GA). Experimental results show that below mentioned GA approaches achieve better performance in terms of scalability and efficiency.