Reproduce “Semantic-Driven Topic Modeling Using Transformer-Based Embeddings and Clustering Algorithms” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Melkamu Abay Mersha, Mesay Gemeda yigezu, Jugal Kalita
The authors present a topic-modeling system that selects labels by measuring candidate words’ relationships to sentences in their document clusters, rather than frequency alone. It uses S-BERT embeddings, UMAP, HDBSCAN clustering, and similarity-based word ranking. On 20 Newsgroups, it reports C_V of 0.735 versus BERTopic’s 0.593. This automated comparison does not establish human interpretability or isolate contextual scoring.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.