HJ-Biplot as a Tool to Give an Extra Analytical Boost for the Latent Dirichlet Assignment (LDA) Model: With an Application to Digital News Analysis about COVID-19

This work objective is to generate an HJ-biplot representation for the content analysis obtained by latent Dirichlet assignment (LDA) of headlines three Spanish newspapers in their web versions referring topic pandemic caused SARS-CoV-2 virus (COVID-19) with more than 500 million affected and almost six deaths date. The used give extra analytical boost model, it easy-to-interpret multivariate technique which does not require in-depth knowledge statistics, allows capturing relationship between topics about COVID-19 news digital newspapers, compares them LDAvis heatmap representations, provides a better visualization, allowing us analyze each newspaper analyzed (column markers represented vectors) 14 from LDA model (row points) plane greatest informative capacity. It concluded that El Mundo 20 M present greater homogeneity published during pandemic, while País presents are less related other two highlighting such as t_12 (Government_Madrid) t_13 (Government_millions).