Topic Modeling on NYT corpus
0.611NPMIECRTM
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| ECRTMCategory=NTMs2025.10 | 0.611 | 0.996 | 3.16 | 2.98 | 3.48 | 20.4 | |
| DecTMCategory=NTMs2025.10 | 0.5542 | 0.988 | 2.86 | 2.9 | 3.42 | 23.6 | |
| TSCTMCategory=NTMs2025.10 | 0.5364 | 0.972 | 2.56 | 2.72 | 2.9 | 20.4 | |
| CombinedTMCategory=NTMs2025.10 | 0.4853 | 0.948 | 3.04 | 2.88 | 3.54 | 23.9 | |
| FASTopicCategory=Others2025.10 | 0.2831 | 0.4973 | 2.44 | 2.38 | 2.75 | 24.8 | |
| BERTopicCategory=Others2025.10 | 0.238 | 0.7605 | 2.97 | 2.06 | 3.08 | 31.4 | |
| Claude-sonnet-4Category=LLMs, Context Window=200k, Zero-shot=true2025.10 | 0.1886 | 0.9845 | 5 | 4.72 | 4.27 | 91.4 | |
| ETMCategory=NTMs2025.10 | 0.1758 | 0.92 | 2.98 | 2.54 | 2.86 | 19.6 | |
| Gemini-3.1Category=LLMs, Context Window=1M, Zero-shot=true2025.10 | 0.1634 | 0.9321 | 4.1 | 3.28 | 4.16 | 93.7 | |
| NSTMCategory=NTMs2025.10 | 0.1509 | 0.744 | 2.8 | 2.58 | 2.84 | 19.2 | |
| Deepseek-v3.1Category=LLMs, Context Window=128k, Zero-shot=true2025.10 | 0.1417 | 0.9439 | 3.46 | 3.53 | 3.93 | 78.3 | |
| Llama-4Category=LLMs, Context Window=1M, Zero-shot=true2025.10 | 0.1403 | 0.8246 | 3.9 | 4 | 3.9 | 69.7 | |
| GPT-5.3Category=LLMs, Context Window=400k, Zero-shot=true2025.10 | 0.1221 | 0.9388 | 5 | 4.93 | 4.07 | 94.4 | |
| Qwen3.5Category=LLMs, Context Window=262k, Zero-shot=true2025.10 | 0.1113 | 0.9667 | 4.5 | 3.83 | 4.16 | 86.2 |