Top-down Span-based Parser with Silver Agreement Subtrees (ensemble)

Reported on 4 benchmarks across 1 task

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

Natural Language Processing4 results

Discourse ParsingonRST-DT
RST-Parseval (Full)· uses extra data
62.6
Discourse ParsingonRST-DT
RST-Parseval (Nuclearity)· uses extra data
75
best: 76 (End-to-end Top-down (XLNet))
Discourse ParsingonRST-DT
RST-Parseval (Relation)· uses extra data
63.2
Discourse ParsingonRST-DT
RST-Parseval (Span)· uses extra data
87.1
best: 87.6 (End-to-end Top-down (XLNet))