“…LDA is a particularly popular parametric approach that models documents as mixtures of topics and topics as mixtures of words (probabilistic distributions over words). An overview of recent papers reporting the application of two widely used LDA-based packages, namely Java-based Mallet 4 [Zhou, Awasthi, Cardinal, 2020;Fang, Partovi, 2021;Cho, Park, Song, 2020] and Python-based Gensim 5 [Porter, 2018;Kastrati, Kurti, Imran, 2020;Riesener et al, 2019], as well as extensive experiments with both packages on the 1st-week data (LJ) followed by the analysis of LDAvis [Sievert, Shirley, 2014] output, we settled on the use of the Mallet package [Mimno et al, 2011]. Ebeid and Arango 6 compare both tools and point out that both have their strengths and weaknesses.…”