Generalized fuzzy hypergraph for link prediction and identification of influencers in dynamic social media networks

Despite the importance of link prediction and identification of influencers in dynamic social media systems, the existing methodical theories are not capable of analyzing complex multilayer relations in social media networks which contain uncertainty. In fact, there is no theoretical exploration concurrently focused on multidimensional and interrelated entities in a fuzzy-based social media environment. To cover this gap, a neoteric generalized fuzzy hypergraph (GFH) methodology is designed using developed n-ary fuzzy relation technique that is the extension of convolutional binary fuzzy relation. Characterizing reflexive, symmetric, transitive, composition, t-cut and support techniques is carried out for multidimensional uncertain-based space. Also, a graphical approach is created in the generalized fuzzy hypergraph to assist the derivation of foundational implications and concepts. The GFH framework can be applied for the intelligent management of complex systems for sole or mass users of local and global social media platforms by adopting specific membership degree for each individual. To predict the linkages between elements, a fuzzy-based indicator FLP (fuzzy link prediction) is promoted, along with the indicator of SIR (score of interaction rate) to identify the influencers (strongest communicators) in an uncertain space. Through the FLP evaluation, the extracted data are analyzed as per the highest value of 1 for single, 3 for binary, 3.8 for triplet, and 0.9 for quaternary spaces for their probable links. Through the analysis of SIR data on the individuals' membership degrees for the usage of social media platforms, the highest interaction value of 0.99 is correlated to a single member, while 5.42 magnitude addresses an influential person. The performance results show that the presented theoretical and structural approach, that is superior to the classical graph theories, is promising to configure intelligent expert systems, predict the likelihood of connections, detect communities, and specify the influencers in real social media platforms that contain uncertainty.

Automated summary

This study proposes a neoteric generalized fuzzy hypergraph (GFH) methodology to analyze complex multilayer relations in social media networks. It introduces a fuzzy link prediction indicator (FLP) and a score of interaction rate (SIR) indicator to predict linkages and identify influencers in an uncertain space. The experimental results demonstrate the potential of this methodology in various areas such as connection prediction, community detection, and influencer identification.