CONDCLOSE: new algorithm of association rules extraction
Hamida Amdouni, Mohamed Mohsen Gammoudi · 2011
Several Methods are proposed for associate rule generation. They try to solve two problems related to redundancy and relevance of associate rules. In this paper we introduce a method called CONDCLOSE which provides the reduction of PRINCE algorithm run-time proposed by Hamrouni and al. in 2005.In fact, we show how the notions of pseudo context and condensed context which we introduce in this paper allows us to attempt these three objectives: no redundancy, relevance of associate rules and minimizing run time of associate rule extraction.