Nowadays, due to the data deluge situation, every computation has to be carried out in voluminous data. The sub-network mining from the complex and voluminous interaction data is one of the research challenges. The highly connected sub-networks will be more cohesive in the network. They are responsible for communication among the network, which is useful for studying their functionalities. A novel score-based co-clustering (MR-CoC) technique with MapReduce is proposed to mine the highly connected sub-network from interaction networks. The MapReduce environment is chosen to cope with complex, voluminous data and to parallelize the computation process. This approach is used to mine cliques, non-cliques, and overlapping sub-network patterns from the adjacency matrix of the network. The complexity of the proposed work is O (Es + log Ns), which is minimal than the existing approaches like MCODE and spectral clustering.
CITATION STYLE
Gowri, R., & Rathipriya, R. (2018). Cohesive sub-network mining in protein interaction networks using score-based co-clustering with mapreduce model (MR-CoC). In Advances in Intelligent Systems and Computing (Vol. 645, pp. 227–236). Springer Verlag. https://doi.org/10.1007/978-981-10-7200-0_20
Mendeley helps you to discover research relevant for your work.