Krahasoni metodat
Shqyrtoni metodat e zgjedhura krah për krah; rreshtat që ndryshojnë janë të theksuar.
| Modelimi i temave me vetë-mbikëqyrje× | Modelimi i temave gjysmë-mbikëqyrës× | |
|---|---|---|
| Fusha | Mësimi i thellë | Mësimi i thellë |
| Familja | Machine learning | Machine learning |
| Viti i origjinës≠ | 2020–2023 | 2009 |
| Krijuesi≠ | Various (Miao et al. 2016 for neural topic models; self-supervised objectives widely adopted 2020–2023) | Ramage, D.; Andrzejewski, D.; and related NLP community |
| Lloji≠ | Self-supervised neural topic model | Probabilistic graphical model (supervised/constrained extension of LDA) |
| Burimi themelues≠ | Wu, X., Li, C., Zhu, Y., & Miao, Y. (2023). Effective Neural Topic Modeling with Embedding Clustering Regularization. Proceedings of the 40th International Conference on Machine Learning (ICML 2023), PMLR 202, 37335–37357. link ↗ | Ramage, D., Hall, D., Nallapati, R., & Manning, C. D. (2009). Labeled LDA: A supervised topic model for credit attribution in multi-labeled corpora. Proceedings of the 2009 Conference on Empirical Methods in Natural Language Processing, 248–256. Association for Computational Linguistics. link ↗ |
| Emërtime të tjera | SSL topic model, self-supervised neural topic model, contrastive topic modeling, self-supervised LM-based topic modeling | semi-supervised LDA, labeled LDA, seed-guided topic modeling, constrained topic model |
| Të lidhura≠ | 5 | 3 |
| Përmbledhja≠ | Self-supervised topic modeling combines the interpretable topic discovery of classical topic models with self-supervised learning objectives — such as contrastive loss, masked language modeling, or reconstruction — to learn coherent, semantically rich topics from unlabeled text without human-annotated labels. It bridges classical probabilistic topic models and modern representation learning, yielding topics better aligned with contextual meaning. | Semi-supervised topic modeling extends unsupervised topic models such as LDA by incorporating partial human supervision — seed words, labeled documents, or must-link/cannot-link constraints — to steer discovered topics toward meaningful, domain-relevant categories while still exploiting the large unlabeled corpus for statistical strength. |
| ScholarGateSeti i të dhënave ↗ |
|
|