Research map / Label-efficient and robust learning / Recognition and long-tail
Jointly Modeling Topics and Intents with Global Order Structure
Bei Chen, Jun Zhu, Nan Yang, Tian Tian, Ming Zhou, Bo Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Add entropic regularization to GMM-LDA (ablation) | GMM-LDA · Chemical and Elements, unsupervised clustering | Fscore: 0.7065 -> 0.7115; 0.6075 -> 0.6339; 0.7017 -> 0.7063; 0.6549 -> 0.6631 | Improved |
| Add entropic regularization to sGMM-LDA (ablation) | sGMM-LDA · Chemical, supervised classification | ACC: 0.731 -> 0.730 | Mixed |
| Add entropic regularization to sGMM-LDA (ablation) | sGMM-LDA · Elements, supervised classification | ACC: 0.752 -> 0.775 | Improved |
| Model topic and intent words jointly instead of using only the intent component (ablation) | GMM-LDA · Chemical, unsupervised clustering, K=5 | ARI: 0.1942 -> 0.4055 | Improved |
| Model topic and intent words jointly instead of using only the intent component (ablation) | GMM-LDA · Elements, unsupervised clustering, K=10 | ARI: 0.3652 -> 0.4013 | Improved |
| Add a non-uniform global intent-order prior instead of assuming uniform intent permutations (ablation) | GMM-LDA · Chemical, unsupervised clustering, K=10 | ARI: 0.3206 -> 0.4477 | Improved |
| Add a non-uniform global intent-order prior instead of assuming uniform intent permutations (ablation) | GMM-LDA · Elements, unsupervised clustering, K=10 | ARI: 0.3432 -> 0.4013 | Improved |
| Use 20% labeled documents during training | GMM-LDA · Chemical, intent prediction | ARI: 0.4055 -> 0.510 | Improved |
| Use 20% labeled documents during training | GMM-LDA · Elements, intent prediction | ARI: 0.4013 -> 0.521 | Improved |
| Use a global order structure rather than a first-order Markov chain | GMM-LDA · Chemical, unsupervised clustering | ARI | Improved |
| Vary the order dispersion parameter across 0.5<rho_0<2 (ablation) | GMM-LDA · Elements, unsupervised clustering, K=10 | F-score | No clear effect |
Closest papers by meaning
- A Statistical Approach for Optimal Topic Model Identification (JMLR 2022 · similarity 0.86)
- E-LDA: Toward Interpretable LDA Topic Models with Strong Guarantees in Logarithmic Parallel Time (ICML 2025 · similarity 0.86)
- textTOvec: DEEP CONTEXTUALIZED NEURAL AUTOREGRESSIVE TOPIC MODELS OF LANGUAGE WITH DISTRIBUTED COMPOSITIONAL PRIOR (ICLR 2019 · similarity 0.85)
- Diagnosing and Improving Topic Models by Analyzing Posterior Variability (AAAI 2018 · similarity 0.85)
- Knowledge-Aware Bayesian Deep Topic Model (NeurIPS 2022 · similarity 0.85)
- Scalable inference of topic evolution via models for latent geometric structures (NeurIPS 2019 · 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.
