Distributed data mining is more efficient, scalable and its performance is better than the central data mining techniques. Incremental DBSCAN algorithm is better than the other method DBSCAN. Incremental DBSCAN can give better performance in distributed environment in terms of run time complexity using Hadoop platform. The proposed system uses spatial dataset i.e. dataset are distributed among different sites. Initially, clusters are generated at each site using DBSCAN algorithm. Then a global cluster is created from these local clusters. Any new data is then either merged with the existing cluster or new cluster is created. Any client request for data is then processed at individual node and response is sent back to the client. The execution time taken by the proposed method is lesser than its predecessors; DBSCAN and IDBSCAN.Keywords: DBSCAN, IDBSCAN, Hadoop, clustering, bigdat Cite this Article: Asha Rajput, Chander Diwakar. Enhanced Incremental DB Scan Algorithm using Map-Reduce. Research & Reviews: A Journal of Embedded System & Applications. 2015; 3(2): 1–6p.
