Abstract
<title>Abstract</title> <p>The process of detecting strongly connected groups of nodes within a network is called Community Detection (CD). Nevertheless, inter and intra-layer dependencies weren’t analyzed in prevailing works, which limits the ability to capture the full extent of community overlap. Thus, inter and intra-layer dependencies-based CD using Skellam Nutcracker Opti-mization Algorithm (S-NOA) and SwishSERF Attention Spatial Drop- Macro Unit BasedConvolutional Neural Network (S2ASD-MUNet) is presented in this work. Primarily, this work takes the input data and constructs the graph network centered on the nodes and edges. Then, it performs the Gower Glorot Dice LeCun Kmeans (G2DL-Kmeans)-based mapping process and S-NOA-based reducing process. After that, the S-NOA-based head node selection from the reduced data followed by node density calculation is done. Meanwhile, the attributes are extracted from the reduced data and given to inter and intra-layer dependencies detection and CD. Similarly, from the selected head nodes, the contextual information is extracted; next, by using S-NOA, inter and intra-layer dependencies are discovered. Lastly, by utilizing the extracted contextual information, inter and intra-layer dependencies output, calculated node density, and extracted attributes, the S2ASD-MUNet-based CD is performed. As per the outcomes, the proposed system attained a high modularity (0.962),thus outperforming the prevailing techniques</p>