방법 비교
선택한 방법을 나란히 검토하세요. 서로 다른 행은 강조 표시됩니다.
| NLP를 위한 대조 학습× | 자기 지도 학습× | |
|---|---|---|
| 분야≠ | 텍스트 마이닝 | 머신러닝 |
| 계열≠ | Process / pipeline | Machine learning |
| 기원 연도≠ | 2020–2021 | 2018–2020 |
| 창시자≠ | Gao, Yao & Chen (SimCSE, 2021); Khosla et al. (Supervised Contrastive, 2020) | LeCun, Y. and community (formalized ~2018–2020) |
| 유형≠ | Self-supervised / supervised representation learning | Representation learning paradigm |
| 원전≠ | Gao, T., Yao, X., & Chen, D. (2021). SimCSE: Simple Contrastive Learning of Sentence Embeddings. Proceedings of EMNLP 2021. link ↗ | LeCun, Y. & Misra, I. (2022). Self-supervised learning: The dark matter of intelligence. Meta AI Blog. https://ai.facebook.com/blog/self-supervised-learning-the-dark-matter-of-intelligence/ link ↗ |
| 별칭 | SimCSE, contrastive sentence embeddings, ContrastiveBERT, Karşıtlık Öğrenmesi — NLP (Contrastive Learning) | SSL, self-supervised pre-training, pretext-task learning, unsupervised representation learning |
| 관련≠ | 4 | 3 |
| 요약≠ | Contrastive learning for NLP is a representation-learning technique — popularised by SimCSE (Gao et al., 2021) and Supervised Contrastive Learning (Khosla et al., 2020) — that trains a text encoder by pulling embeddings of similar text pairs together while pushing embeddings of dissimilar pairs apart. The result is a dense, high-quality embedding space that can be learned with no labels at all, or with minimal supervision, making it especially valuable when annotated data are scarce. | Self-supervised learning (SSL) is a machine-learning paradigm that generates its own supervisory signal directly from unlabeled data by defining an auxiliary pretext task — such as predicting masked words, rotating images, or contrasting augmented views — and uses the learned representations as a powerful starting point for downstream tasks with minimal labeled examples. |
| ScholarGate데이터셋 ↗ |
|
|