A Novel Scheme for Efficiency of Customer Loyalties Using Pattern Growth Method with Upper Bound Algorithm

Abstract

The aim is to improve and develop a novel scheme to detect loyalties of customers using pattern growth method.Novel Pattern growth method compared with upper bound taxonomy sequence algorithm are used to detect online sales customer loyalties. Sample size is determined using the G Power calculator and found to be 10 per group. Totally 20 samples are used. Pretest power is 80% with CI of 95%. Based on the analysis Novel Pattern growth method has an accuracy of 80.8% and upper bound taxonomy sequence algorithm has 67.25%. Significance value is 0.0001 (p<0.05, two-tailed). Proposed algorithm Novel Pattern-Growth method has higher accuracy than Upper Bound Taxonomy for selected datasets for more reviews.