BioSentVec (PubMed + MIMIC-III)

Reported on 4 benchmarks across 2 tasks · 1 paper · 2 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing2 results

Sentence EmbeddingsonMedSTS
Pearson Correlation· uses extra data· 2018-10-22
0.767
SOTA
BioSentVec: creating sentence embeddings for biomedical texts arXiv:1810.09302
Sentence EmbeddingsonBIOSSES
Pearson Correlation· uses extra data· 2018-10-22
0.795
best: 0.871 (Supervised combination of: Jaccard, Q-gram, sent2vec, Paragraph vector DM, skip-thoughts, fastText)
BioSentVec: creating sentence embeddings for biomedical texts arXiv:1810.09302

Methodology2 results

Representation LearningonMedSTS
Pearson Correlation· uses extra data· 2018-10-22
0.767
SOTA
BioSentVec: creating sentence embeddings for biomedical texts arXiv:1810.09302
Representation LearningonBIOSSES
Pearson Correlation· uses extra data· 2018-10-22
0.795
best: 0.871 (Supervised combination of: Jaccard, Q-gram, sent2vec, Paragraph vector DM, skip-thoughts, fastText)
BioSentVec: creating sentence embeddings for biomedical texts arXiv:1810.09302