Semantic Role Labeling

Semantic role labeling aims to model the predicate-argument structure of a sentence and is often described as answering "Who did what to whom". BIO notation is typically used for semantic role labeling.

Example:

| Housing | starts | are | expected | to | quicken | a | bit | from | August’s | pace | | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | | B-ARG1 | I-ARG1 | O | O | O | V | B-ARG2 | I-ARG2 | B-ARG3 | I-ARG3 | I-ARG3 |

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 391–400 of 620 papers

Title	Date	Tasks	Status
Memory-Based Acquisition of Argument Structures and its Application to Implicit Role Detection	Sep 1, 2015	Natural Language InferenceQuestion Answering	—Unverified
Identifying Key Concepts from EHR Notes Using Domain Adaptation	Sep 1, 2015	Domain AdaptationInformation Retrieval	—Unverified
Semantic Framework for Comparison Structures in Natural Language	Sep 1, 2015	Open-Domain Question AnsweringQuestion Answering	—Unverified
Any-language frame-semantic parsing	Sep 1, 2015	Cross-Lingual TransferQuestion Answering	—Unverified
Question-Answer Driven Semantic Role Labeling: Using Natural Language to Annotate Natural Language	Sep 1, 2015	Semantic Role Labeling	—Unverified
Hierarchical Recurrent Neural Network for Document Modeling	Sep 1, 2015	ChunkingLanguage Modelling	—Unverified
Distributed Representations for Unsupervised Semantic Role Labeling	Sep 1, 2015	Language ModellingMachine Translation	—Unverified
Chinese Semantic Role Labeling with Bidirectional Recurrent Neural Networks	Sep 1, 2015	Chinese Semantic Role LabelingFeature Engineering	—Unverified
Semantic Role Labeling with Neural Network Factors	Sep 1, 2015	Semantic Role LabelingSemantic Textual Similarity	—Unverified
Joint A* CCG Parsing and Semantic Role Labelling	Sep 1, 2015	Semantic ParsingSemantic Role Labeling	—Unverified

Show:10 25 50

← PrevPage 40 of 62Next →

All datasets OntoNotes CoNLL 2005 CoNLL05 Brown CoNLL05 WSJ CoNLL12 CoNLL-2009 CoNLL-2012

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	HeSyFu	F1	88.59	—	Unverified
2	CRF2o + RoBERTa	F1	88.32	—	Unverified
3	MRC-SRL	F1	88.3	—	Unverified
4	ReCAT(pretrained on wikitext103)	F1	88	—	Unverified
5	SRL-MM + XLNet	F1	87.67	—	Unverified
6	CRF2o + BERT	F1	87.66	—	Unverified
7	RoBERTa+RegCCRF	F1	87.51	—	Unverified
8	RoBERTa+CRF	F1	87.27	—	Unverified
9	BiLSTM-Span (Ensemble)	F1	87	—	Unverified
10	BiLSTM-Span	F1	86.2	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	MRC-SRL	F1	90	—	Unverified
2	SRL-MM + XLNet	F1	89.8	—	Unverified
3	CRF2o + RoBERTa	F1	89.54	—	Unverified
4	HeSyFu	F1	89.04	—	Unverified
5	CRF2o + BERT	F1	89.03	—	Unverified
6	Mohammadshahi and Henderson (2021)	F1	88.93	—	Unverified
7	BiLSTM-Span (Ensemble, predicates given)	F1	88.5	—	Unverified
8	CRF2o	F1	87.87	—	Unverified
9	Li et al. (2019) (Ensemble)	F1	87.7	—	Unverified
10	BiLSTM-Span	F1	87.6	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	DeepStruct multi-task w/ finetune	F1	92.1	—	Unverified
2	DeepStruct multi-task	F1	92	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	DeepStruct multi-task	F1	95.5	—	Unverified
2	DeepStruct multi-task w/ finetune	F1	95.2	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	DeepStruct multi-task	F1	97.2	—	Unverified
2	DeepStruct multi-task w/ finetune	F1	96	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	Ours (High-Order model)	F1 (Arg.)	90.2	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	HeSyFu	Avg. F1	88.59	—	Unverified