Papers › Lexical Semantic Recognition

Lexical Semantic Recognition

30 Apr 2020ACL (MWE) 2021 8arXiv:2004.15008archive 2025-07-28

Nelson F. Liu, Daniel Hershcovich, Michael Kranzlein, Nathan Schneider

In lexical semantics, full-sentence segmentation and segment labeling of various phenomena are generally treated separately, despite their interdependence. We hypothesize that a unified lexical semantic recognition task is an effective way to encapsulate previously disparate styles of annotation, including multiword expression identification / classification and supersense tagging. Using the STREUSLE corpus, we train a neural CRF sequence tagger and evaluate its performance along various axes of annotation. As the label set generalizes that of previous tasks (PARSEME, DiMSUM), we additionally evaluate how well the model generalizes to those test sets, finding that it approaches or surpasses existing models despite training only on STREUSLE. Our work also establishes baseline models and evaluation metrics for integrated and accurate modeling of lexical semantics, facilitating future work in this area.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

nert-nlp/streusle officialmentioned in papermentioned on GitHub report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Natural Language UnderstandingSentenceSentence segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Natural Language Understanding STREUSLE BERT (pred POS/lemmas) Full F1 (Preps) 71.6 #1 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (pred POS/lemmas) Function F1 (Preps) 82.8 #1 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (pred POS/lemmas) Role F1 (Preps) 72.4 #1 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (pred POS/lemmas) Tags (Full) Acc 82.5 #1 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (none) Full F1 (Preps) 70.9 #2 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (none) Function F1 (Preps) 81.0 #2 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (none) Role F1 (Preps) 71.9 #2 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (none) Tags (Full) Acc 82.0 #2 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (gold POS/lemmas) Full F1 (Preps) 71.4 #3 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (gold POS/lemmas) Function F1 (Preps) 81.7 #3 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (gold POS/lemmas) Role F1 (Preps) 72.4 #3 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE BERT (gold POS/lemmas) Tags (Full) Acc 81.0 #3 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (gold POS/lemmas) Full F1 (Preps) 61.0 #4 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (gold POS/lemmas) Tags (Full) Acc 79.3 #4 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (none) Full F1 (Preps) 58.1 #5 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (none) Tags (Full) Acc 77.5 #5 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (pred POS/lemmas) Full F1 (Preps) 58.0 #6 of 11 Archive leaderboard report
Natural Language Understanding STREUSLE GloVe (pred POS/lemmas) Tags (Full) Acc 77.1 #6 of 11 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

CRF

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections