Replication Data for: L2R algorithm for Likelihood Estimation in GaP-FA

This repository contains the document collections and the trained GaPs for reproducing experiments in "L2R algorithm for Likelihood Estimation in GaP-FA". In particular, the repository is composed of: - DATASETNAME_*d_*w.dat files which contain the processed dataset. - DATASETNAME_dictionary.txt which contain the vocabulary. - DATASETNAME_GaP_*N_*W_100T.mat which contain the GaP models trained in the complete dataset with a maximum of 100 topics. - DATASETNAME_GaP_*N_*W_5T.mat which contain the GaP models trained in the downsided dataset with a maximum of 5 topics.