Abstract
Association rule mining is a one of the most important technique in data mining. It extracts significant patterns from transaction databases and generates rules used in many decision support application. Modern organizations are geographically distributed. Using the traditional centralized association rule mining to discover useful patterns in such distributed system is not always feasible because merging data sets from different sites into a centralized site incurs huge network communication and time costs. This paper present an optimized Distributed Association Rule Mining (D-ARM) based on vertical partitioning. The existing D-ARM algorithms have lots of communication overhead, which is a major issue for concerning. The proposed approach minimizes this communication overhead and it is based on total count. The papers then discuss the Total Count on Vertical Dataset (TCDV) use of this structure which offers significant advantages with respect to existing DARM techniques.
