Spark-Based FP-Growth Algorithm for Generating Association Rules from Big Data

Abstract

AbstractData mining is used to discover Business Intelligence Rules from large transactional database, frequent itemset mining and association rule mining are data mining techniques which are utilized for real-world applications using apriori algorithm as well as FP-Growth algorithm. To extract frequent itemset and association rules from the grocery store using traditional apriori algorithm takes time because it generates candidate key for each item in the dataset, whereas in FP-Tree without generating any candidate key it finds frequent items and association rules for the grocery dataset by constructing FP-Tree, and it is time consuming. To overcome this problem, we propose a method called MapReduce-based FP-Tree algorithm which generates frequent patterns and association rules by using parallel computations to reduce computational time. The experimental results show that time taken for generating frequent patterns and association rules for the grocery dataset is less.KeywordsAssociation rulesBig dataData analyticsFP-growthFrequent patternsHadoopMap reduce