Table 8

Ranker optimization on BSP: Showing ranker performance for various feature type combinations compared with a random-choice baseline, only considering instances where the gold-standard scope aligns to a constituent within the 1-best parse.

Features
F1
Baseline
26.76
Path 78.10 
Path+Surface 79.93 
Path+Linguistic 83.72 
Path+Surface+Linguistic 85.30 
Features
F1
Baseline
26.76
Path 78.10 
Path+Surface 79.93 
Path+Linguistic 83.72 
Path+Surface+Linguistic 85.30 
Close Modal

or Create an Account

Close Modal
Close Modal