Mining Association Rules from Stream Data Based on the Dynamic Support

Abstract

The Stream data exists in the field of industrial production, life activities, business transactions, and other fields. It is closely related to people's life, production and so on. This paper proposes inter-transaction association rules mining method based on dynamic support threshold. Inter-transaction association rules refer to the association rules between different time periods. This paper firstly uses the sliding window to limit stream data, then do preprocessing on stream data. In the process of pretreatment using linearization method fitting to raw data and it reduce the amount of data at the same time, and finally at the end of the preprocessing, generating large transaction grouping method of inter transaction association rules is proposed in this paper. This paper uses conceptual data attenuation, thereby reducing the influence of old data to the mining result. Due to artificial setting minimum support threshold may bring many problems, so this paper presents a method for searching minimum support threshold.