Usporedite metode

	Prenosno učenje sa semantičkim ugradnjama	Prenosno učenje s klasifikacijom utemeljenom na BERT-u
Područje	Duboko učenje	Duboko učenje
Obitelj	Machine learning	Machine learning
Godina nastanka	2017–2019	2019 (BERT); transfer learning paradigm established circa 2010
Tvorac	Reimers, N. & Gurevych, I. (SBERT); Conneau, A. et al. (InferSent)	Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (BERT); Pan, S. J. & Yang, Q. (transfer learning survey)
Vrsta	Transfer learning / sentence representation	Pre-trained transformer fine-tuned for classification
Temeljni izvor	Reimers, N. & Gurevych, I. (2019). Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing (EMNLP), 3982–3992.	Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2019). BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. In Proceedings of NAACL-HLT 2019, 4171–4186. Association for Computational Linguistics.
Drugi nazivi	sentence embedding transfer learning, pre-trained sentence encoder fine-tuning, SBERT transfer learning, sentence representation transfer	BERT fine-tuning for classification, BERT transfer learning classifier, pre-trained BERT classifier, BERT downstream classification
Srodne	5	4
Sažetak	Transfer Learning with Sentence Embeddings takes a large pre-trained encoder — such as Sentence-BERT or the Universal Sentence Encoder — that already encodes general language knowledge into fixed-length vectors, and adapts it to a new task or domain with little additional labelled data. The pre-trained representations give a head start that often outperforms task-specific models trained from scratch on modest corpora.	Transfer Learning with BERT-based Classification adapts a large transformer language model, pre-trained on massive text corpora, to a target classification task by fine-tuning its weights on labeled examples. The pre-trained representations encode rich syntactic and semantic knowledge, enabling high accuracy even when the labeled dataset is small.