Incremental parallelization clustering method of massive network data based on DBSCAN algorithm
Traditional clustering algorithms require running the entire clustering process again when facing dynamically in-creasing data,which is time-consuming and inefficient.To effectively address this challenge,a massive network data incremental parallelization clustering method based on DBSCAN algorithm is proposed.Using the Chernoff bounds criterion to partition network data,ensuring balance and representativeness.Applying the DBSCAN algorithm for clustering,accurately identifying high-density areas,while processing noisy data,to achieve initial clustering of network data.For dynamic data,set the principle of incremental merging to efficiently merge new data with original clusters and maintain real-time updates of clustering results.The experimental results show that the proposed method has a high confidence level(not less than 97%)and performs well in clustering time com-plexity,successfully achieving incremental parallelization,precise and fast clustering of massive network data.
万方数据
