Topic Modeling on NCTBText benchmark leaderboard · SOTA2 Research

Topic Modeling on NCTBText

0.87CVGHTM

Evaluation Results

Method
GHTM2025.08	0.87	0.27	0.99	1
BERTopic (SBERT+UMAP+HDBSCAN)2025.08	0.82	-0.08	0.44	0.52
BERTopic (SBERT+PCA+HDBSCAN)2025.08	0.78	-0.05	0.49	0.61
BERTopic (SBERT+SVD+HDBSCAN)2025.08	0.78	-0.04	0.51	0.62
BERTopic (SBERT+HDBSCAN)2025.08	0.77	-0.04	0.55	0.74
Top2vec (Doc2Vec)2025.08	0.74	0.11	1	1
GCTM2025.08	0.73	0.05	1	1
BERTopic (SBERT+UMAP+KMeans)2025.08	0.69	0.02	0.97	0.99
ProdLDA2025.08	0.66	0.05	1	1
BERTopic (SBERT+UMAP+Spectral)2025.08	0.66	0.01	0.93	0.98
BERTopic (SBERT+UMAP+DBSCAN)2025.08	0.65	-0.09	0.76	0.86
Combined TM (BOW+SBERT)2025.08	0.64	0.05	1	1
ZeroShot TM (BOW+SBERT)2025.08	0.64	0.05	1	1
Graph2Topic2025.08	0.59	0.08	0.69	0.68
BERTopic (SBERT+UMAP+Agglomerative)2025.08	0.58	-0.09	0.86	0.96
NMF (TF-IDF)2025.08	0.57	0.04	0.95	0.99
BERTopic (GloVe+UMAP+HDBSCAN)2025.08	0.56	0.13	0.68	0.91
GNTM2025.08	0.56	0.04	0.89	0.97
NMF (BOW)2025.08	0.54	0.06	0.82	0.96
Neural LDA2025.08	0.54	-0.3	1	1
CluWords2025.08	0.52	-0.06	0.9	1
LDA (TF-IDF)2025.08	0.51	-0.04	0.96	0.99
LDA (BOW)2025.08	0.5	0.03	0.83	0.95
BERTopic (Word2Vec+UMAP+HDBSCAN)2025.08	0.44	-0.05	0.73	0.9
BERTopic (Doc2Vec+UMAP+HDBSCAN)2025.08	0.44	-0.05	0.72	0.9
LSA (TF-IDF)2025.08	0.42	-0.07	0.61	0.78
BERTopic (Fast Text+UMAP+HDBSCAN)2025.08	0.42	-0.07	0.79	0.93
ETM (FastText)2025.08	0.4	-0.14	0.81	0.94
LSA (BOW)2025.08	0.34	-0.05	0.47	0.78