Research map / Graph learning / Knowledge and temporal graphs
TopicNet: Semantic Graph-Guided Topic Discovery
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Represent topic embeddings as Gaussian distributions rather than deterministic vectors (ablation) | SawETM · 20NG, RCV1, and Wiki | per-heldout-word perplexity: 20NG: 688 -> 685 (depth 5); RCV1: 873 -> 866 (depth 5); Wiki: 1503 -> 1477 (depth 5) | Helped |
| Represent topic embeddings as Gaussian distributions rather than deterministic vectors (ablation) | SawETM · 20NG | classification accuracy: 71.1±0.2 -> 75.2±0.4 | Helped |
| Add WordNet hierarchical prior regularization (ablation) | Gaussian SawETM · 20NG and R8, vocabulary size 20,000 | classification accuracy: 20NG: 75.2±0.4 -> 74.3±0.5; R8: 91.0±0.3 -> 89.5±0.4 | Hurt |
| Add WordNet hierarchical prior regularization (ablation) | Gaussian SawETM · 20NG and R8, vocabulary size 20,000 | clustering ACC: 20NG: 53.0±0.3 -> 50.8± 0.3; R8: 65.2±0.4 -> 64.2±0.3 | Hurt |
| Add WordNet hierarchical prior regularization (ablation) | Gaussian SawETM · 20NG and R8, vocabulary size 20,000 | clustering NMI: 20NG: 51.3±0.2 -> 47.8±0.4; R8: 47.0±0.3 -> 46.5±0.2 | Hurt |
| Use a better-fitted WordNet semantic graph for 20NG (ablation) | Gaussian SawETM · 20NG, vocabulary size 2,000 | classification accuracy: 62.8±0.3 -> 65.0±0.3 | Helped |
Closest papers by meaning
- Sawtooth Factorial Topic Embeddings Guided Gamma Belief Network (ICML 2021 · similarity 0.92)
- Inter and Intra Topic Structure Learning with Word Embeddings (ICML 2018 · similarity 0.90)
- Knowledge-Aware Bayesian Deep Topic Model (NeurIPS 2022 · similarity 0.90)
- Bayesian Deep Embedding Topic Meta-Learner (ICML 2022 · similarity 0.89)
- On the Affinity, Rationality, and Diversity of Hierarchical Topic Modeling (AAAI 2024 · similarity 0.88)
- FASTopic: Pretrained Transformer is a Fast, Adaptive, Stable, and Transferable Topic Model (NeurIPS 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.