Top-n Utility Patterns Mining Algorithm In Dynamic Databases

With the advent of the era of big data,all works of life are faced with a large amount of data to be processed.How to extract valuable information quickly and efficiently has become an important research in the field of data mining.Utility patterns mining is an important branch of big data.Based on frequent pattern mining,utility patterns mining is considered to make up for the defects of frequent pattern mining.At present,utility patterns mining algorithm has gradually matured.This paper introduces the problem of utility patterns mining and its branching problems,and introduces and compares the existing algorithms of utility patterns mining and its branching problems.Considering the frequent use of dynamic database and the user's demand for top-n utility patterns,this paper proposes a dynamic database top-n utility patterns mining problem,and proposes TOPNTD algorithm for solving this problem.In this paper,the data structure,pruning strategy and optimization strategy of TOPNTD algorithm are designed.The innovative short list structure,pruning strategy and subordinate tagging strategy for dynamic database top-n utility patterns mining problem are proposed.In this paper,the TOPNTD algorithm is implemented by programming and tested in the big data set.Compared to similar algorithms,TOPNTD algorithm is better than rebooting statical top-n utility patterns mining algorithms when solving the dynamic database top-n utility patterns mining problem.