Publication

Abstract : Multi-attribute networks are complex systems that can record connections between items based on a range of qualities.These networks have a wide range of applications, including link prediction, anomaly detection, and community discovery.The primary goal of this research is to identify relevant nodes in a multilayered attributed network. To do this, clustering techniques based on numerous attributes over multiple layers are used. The attributes taken into consideration are H - Index, P - Index, Unequal P - Index, and affiliations. The nodes of the citation network were grouped using K-means based on their affiliation similarity scores. The most influential nodes were then identified by an examination of the clusters formed by K-means. We formulated combined attribute score based on attributes, for which DBSCAN is applied across several layers and influential nodes were discovered were developed. Multiple centralities analysis is performed on the clusters found by K-means and DBSCAN. The findings demonstrate that the multilayer network can be utilised to analyse impact and contagion in real-world citation networks.