Abstract
The Hadoop distributed file system offers efficient Mapreduce frame work using which the big datasets can be processed with efficient time complexity. Capability to load on low-cost commodity hardware and greater extent of fault tolerance leading many business organizations to store data in Hadoop distributed file system. Considering the real-time importance of distributed file system in recent literature conventional data mining algorithms getting extended to scale in MapReduce architecture. In line to this trend we propose a fuzzy associative classification algorithm based on MapReduce framework to extract intuitive classification rules from data stored in distributed file systems. The experimental investigation shows that the proposed algorithm on MapReduce frame work can scale to effectively extract intuitive classification rules from training data stored in distributed file systems.