Research map / Label-efficient and robust learning / Semi-supervised learning
Learning Supervised Topic Models for Classification and Regression from Crowds
Filipe Rodrigues, Mariana Lourenço, Bernardete Ribeiro, Francisco C. Pereira
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 |
|---|---|---|---|
| use the multi-annotator model MA-sLDAc instead of LDA+LogReg, Raykar, Rodrigues, Blei 2003, sLDA with majority voting | MA-sLDAc · 20 Newsgroups (simulated annotators) | accuracy | Improved |
| use MA-sLDAc instead of baselines | MA-sLDAc · Reuters-21578 (Mechanical Turk) | accuracy | Improved |
| use MA-sLDAc instead of baselines incl. Bosch 2006 | MA-sLDAc · LabelMe (Mechanical Turk) | accuracy | Improved |
| use stochastic variational inference instead of batch | MA-sLDAc · 20 Newsgroups | log marginal likelihood | Improved |
| use batch instead of svi inference | MA-sLDAc · Reuters-21578 and LabelMe | accuracy | Mixed |
| model annotator noise and bias (multi-annotator baselines) instead of majority voting | LDA + Raykar / Rodrigues · Reuters-21578 and LabelMe | accuracy | Improved |
| use MA-sLDAr instead of LDA+LinReg and sLDA with mean answers | MA-sLDAr · we8there (simulated annotators) | R2 | Improved |
| use MA-sLDAr instead of baselines | MA-sLDAr · movie reviews (Mechanical Turk) | R2 | Improved |
| aggregate answers by majority voting versus the average worker | majority voting · Reuters-21578 | accuracy: 0.568 -> 0.710 | Improved |
| model per-annotator bias parameter | MA-sLDAr · movie reviews | bias estimation | Improved |
Closest papers by meaning
- Algorithms for Generalized Topic Modeling (AAAI 2018 · similarity 0.87)
- Provable Algorithms for Inference in Topic Models (ICML 2016 · similarity 0.86)
- Neural Dynamic Focused Topic Model (AAAI 2023 · similarity 0.86)
- Discovering Discrete Latent Topics with Neural Variational Inference (ICML 2017 · similarity 0.86)
- A Deep and Autoregressive Approach for Topic Modeling of Multimodal Data (TPAMI 2016 · similarity 0.86)
- Coupled Confusion Correction: Learning from Crowds with Sparse Annotations (AAAI 2024 · similarity 0.84)
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.
