Fast and Accurate Neural CRF Constituency Parsing

Estimating probability distribution is one of the core issues in the NLP field. However, in both deep learning (DL) and pre-DL eras, unlike the vast applications of linear-chain CRF in sequence labeling tasks, very few works have applied tree-structure CRF to constituency parsing, mainly due to the complexity and inefficiency of the inside-outside algorithm... (read more)

PDF Abstract IJCAI 2020 PDF IJCAI 2020 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Constituency Parsing CTB5 CRF Parser F1 score 89.80 # 2
Constituency Parsing CTB5 CRF Parser + BERT F1 score 92.27 # 1
Constituency Parsing CTB7 CRF Parser F1 score 88.60 # 2
Constituency Parsing CTB7 CRF Parser + BERT F1 score 91.55 # 1
Constituency Parsing Penn Treebank CRF Parser F1 score 94.12 # 9
Constituency Parsing Penn Treebank CRF Parser + BERT F1 score 95.69 # 1

Methods used in the Paper