Research on Mining Sequential Positive and Negative Association Rules

Mining sequential positive and negative association rules is to mine the inner association or the causal relationship among data in sequential database, which will find some rules that have practical significance for the industry decision-making analysis among the sequence. This paper proposes the relational notions of sequential positive and negative association rule. Based on the new questions when mining the positive and negative rules in the sequential database, the paper discusses the solutions and proposes an algorithm called SPNARM to mine sequential positive and negative association rules (SPNAR). Example analysis results show that SPNARM algorithm is more efficient for mining SPNARs.

I. Introduction

Sequential pattern mining (SPM) has been an important topic of data mining research, and has a wide range of applications, e.g., analysis of DNA sequence patterns, customer purchase behavior analysis, network access mode of analysis, the analysis of scientific experiments, disease treatment early diagnosis, and prediction of natural disasters and so on. The task of sequential pattern mining is to identify all sequence database support which is not less than the minimum support threshold of frequent sequences (sequence mode), It is firstly proposed by Agrawal R. et al. in the shopping basket data analysis [10]. In the past decades, many algorithms on SPM have been proposed. For example Srikant et al. proposed the GSP method in [11], Zaki proposed the SPADE method in [12]. In addition, Constraint-based sequential pattern mining algorithm, based on the pattern of growth sequential pattern mining methods, and databases based on the projection of the sequential pattern mining methods have been proposed. And moreover, there are some expansions of research on SPM, such as closed sequential pattern mining, parallel mining, distributed mining, multi-dimensional sequential pattern mining and approximate sequential pattern mining.