LDAShiny - Interactive Topic Modeling and Bibliometric Analysis via Shiny
Provides a 'Shiny' graphical interface for the complete
workflow of Latent Dirichlet Allocation (LDA) topic modelling
on bibliometric data from Scopus and Web of Science. Steps
include data import and deduplication, text preprocessing
(stopword removal, stemming, n-grams, sparse-term filtering),
statistical inference to select the optimal number of topics
via coherence, final model training, and topic trend analysis
over time using linear regression. All results can be exported
as Excel files, RDS objects, and publication-quality plots.
