Influence Maximization (IM) is a problem of detecting a small set of highly influential users in a social network. Application areas of IM include the spread of news, viral marketing, and the outbreak of diseases. In most of the existing IM approaches, the nodes' structural information has been considered for computing their influence spread ability. The users' interest, their interaction behavior, popularity, and location sharing information are being neglected. Although many existing works have considered a few of these measures; however, they do not consider them collectively and face challenges in time efficiency and suitable seed node accuracy. To overcome these challenges, this paper proposes Context-Aware Influential Nodes Tracking (CINT) algorithm, which uses users' interest, popularity, location information to compute the topic-based diffusion ability of the network, find the topic-wise influential seeds and finally, evaluate the spread of influence for a given message/product. We propose a Contextaware Independent Cascade (CIC) model and a Topic-aware Influence sub-Graph (TIG) model to make our framework efficient and effective. Experimental results on six real-world networks show that the proposed model performs better in terms of effective influence spread as compared to the considered existing state-of-the-art influence maximization algorithms.
Singh, Ashwini Kumar; Kailasam, Lakshmanan; Pradhan, Tribikram; and Gupta, Deepti, "Context-Aware Influential Nodes Tracking in Online Social Networks" (2021). PACIS 2021 Proceedings. 177.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.