Natural Language Understanding On Streusle
评估指标
Full F1 (Preps)
评测结果
各个模型在此基准测试上的表现结果
模型名称 | Full F1 (Preps) | Paper Title | Repository |
---|---|---|---|
RoBERTa + Linear | 78.2 | - | - |
GloVe (pred POS/lemmas) | 58.0 | Lexical Semantic Recognition | |
BERT (gold POS/lemmas) | 71.4 | Lexical Semantic Recognition | |
BiLSTM + MLP (auto syntax) | 53.6 | Comprehensive Supersense Disambiguation of English Prepositions and Possessives | |
GloVe (none) | 58.1 | Lexical Semantic Recognition | |
BERT (pred POS/lemmas) | 71.6 | Lexical Semantic Recognition | |
SVM (feature-rich, gold syntax) | 59.5 | Comprehensive Supersense Disambiguation of English Prepositions and Possessives | |
BiLSTM + MLP (gold syntax) | 58.9 | Comprehensive Supersense Disambiguation of English Prepositions and Possessives | |
GloVe (gold POS/lemmas) | 61.0 | Lexical Semantic Recognition | |
SVM (feature-rich, auto syntax) | 55.7 | Comprehensive Supersense Disambiguation of English Prepositions and Possessives | |
BERT (none) | 70.9 | Lexical Semantic Recognition |
0 of 11 row(s) selected.