Disease and drug network structure in link prediction

Abstract

Due to the growing number of users, social network analysis faces several challenges, including evolution, community detection, and link prediction. Among these issues, link prediction appears to receive the most attention. It is difficult to predict when a new connection might be possible within the destiny. Presently, link prediction techniques can be classified as learning-based or similarity-based. Transferring data records entirely to digital platforms has resulted in data becoming embodied and quantifiable. This study initially established disease and drug networks to track the interaction between diseases and drugs. These networks are comprised of disease diagnosis and physician-written medications. Following the establishment of disease and drug networks, the similarity of nodes was used to predict the connections. The experimental results indicate that the proposed method produces acceptable results.

Keywords

• Social network analysis
• Link prediction
• Disease
• And drug networks
• Machine learning
• Storage