期刊名称：Journal of Emerging Trends in Computing and Information Sciences
电子版ISSN：2079-8407
出版年度：2016
卷号：7
期号：2
页码：48-54
出版社：ARPN Publishers
摘要：The most important step in building any classifier in data mining is to learn the association rules from any given transactions database. Apriori Algorithm is considered to be one of the most popular algorithms, used to find the frequent item sets from a given transactions database. These frequent item sets form the base for building the association rules which are the backbone for the classifiers. The traditional Apriori algorithm is dealt with static transaction database that means the frequent item sets is fixed and as a consequent, the association rules are also fixed. So if any change happened to the transactions database, we must repeat the implementation of the Apriori algorithm in addition to the building the classifier. In this paper we propose a new approach which let the classifier to deal with dynamic transactions database. In our approach, the basic idea is to analyze the effect of this new transaction on the frequent item set and the association rules. The first step in our new approach is to implement the Apriori algorithm and then to complete the next steps to build the classifier. The new feature of our approach is the ability to deal with any new transaction which may be added to the transaction database.
关键词：Classification; data mining; association rules; apriori algorithm.
