A knowledge discovery framework for COVID-19 disease from PubMed abstract using association rule hypergraph

Coronavirus disease 2019 (COVID-19) infects the human being through ACE2 receptors. About 66,94,59,397 humans worldwide had been affected with coronavirus until January 2023. This chapter mainly focuses to extract trends and their association related to the COVID-19 from PubMed documents. The association rule hypergraph is proposed to identify the multiway relationship between genes, side effects, and proteins in PubMed connected to COVID-19 disorders. The words in the abstract that have been preprocessed using Natural Language ToolKit (NLTK) are considered to be vertices in the construction of the hypergraph, which uses the abstract as an edge. Association Rule Mining algorithm is applied on the hypergraph, and the association between the terms is identified. It detects the effective keynotes about the COVID-19 with proper support. Our substructure enabled us to track the evolution of trends in COVID-19 research and the important in demand areas of interest (hotspots) on the pandemic.