TACG: A Topic-Aware Community-Guided Framework for Context-Aware Influence Maximization

Influence maximization (IM) aims to identify a small set of influential nodes that can maximize the propagation of information in a social network. Many studies have selected seed nodes based on the network’s structural features, which may identify inactive spreaders in a particular context. A limited work finds the topic-aware seed set from the network. Among them, a recent study, the TAICM (Topic-Aware Information Coverage Maximization)framework, identifies a seed set (RSS-Degree, RSS-SG algorithms) that can maximize information coverage using a greedy approach under the Independent Cascade model. However, we have examined the computational time of TAICM, which is enormous due to the consideration of multihop information in the calculation of the seed set, and it may find seeds from the same region of the network. To address these limitations, this paper proposes TACG (Topic-Aware Community-based Greedy) approach. This framework identifies the seed set from non-overlapping regions to maximize the spreading across the network. In addition, to reduce the time complexity of finding the seed set of various topics, the proposed TACG used a three-hop neighborhood topic-based feature of a node. Experimental evaluations on two real-world networks, such as the Facebook social network and the Cit-HepPh networks, demonstrate that TACG achieves higher or comparable information coverage compared to existing topic-aware influence maximization algorithms, like RSS-SG, RSS-Degree, and TA-IMRank. TACG achieves up to [7]–[12] % higher coverage than the others while generating seed sets that clearly reflect topic relevance.