Papers › Methods for Recognizing Nested Terms

Methods for Recognizing Nested Terms

22 Apr 2025arXiv:2504.16007archive 2025-07-28

Igor Rozhkov, Natalia Loukachevitch

In this paper, we describe our participation in the RuTermEval competition devoted to extracting nested terms. We apply the Binder model, which was previously successfully applied to the recognition of nested named entities, to extract nested terms. We obtained the best results of term recognition in all three tracks of the RuTermEval competition. In addition, we study the new task of recognition of nested terms from flat training data annotated with terms without nestedness. We can conclude that several approaches we proposed in this work are viable enough to retrieve nested terms effectively without nested labeling of them.

PaperPDFCode

Code

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

Dialogue EvaluationNERNamed Entity RecognitionNamed Entity Recognition (NER)Nested Named Entity RecognitionNested Term ExtractionNested Term RecognitionNested Term Recognition from Flat SupervisionTerm Extractionnamed-entity-recognition

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Nested Term Extraction RuTermEval (Track 1) full nested Scoreboard F1 0.7940 #1 of 1 Archive leaderboard report
Nested Term Extraction RuTermEval (Track 2) full nested Scoreboard Class-agnostic F1 0.78 #1 of 1 Archive leaderboard report
Nested Term Extraction RuTermEval (Track 2) full nested Scoreboard Weighted F1 0.6997 #1 of 1 Archive leaderboard report
Nested Term Extraction RuTermEval (Track 3) full nested Scoreboard Class-agnostic F1 0.60 #1 of 1 Archive leaderboard report
Nested Term Extraction RuTermEval (Track 3) full nested Scoreboard Weighted F1 0.4823 #1 of 1 Archive leaderboard report
Nested Term Recognition from Flat Supervision RuTermEval (Track 1) lemm. inc. + early dmg Scoreboard F1 0.7281 #1 of 1 Archive leaderboard report
Nested Term Recognition from Flat Supervision RuTermEval (Track 2) lemm. inc. + early dmg Scoreboard Class-agnostic F1 0.7337 #1 of 1 Archive leaderboard report
Nested Term Recognition from Flat Supervision RuTermEval (Track 2) lemm. inc. + early dmg Scoreboard Weighted F1 0.631 #1 of 1 Archive leaderboard report
Nested Term Recognition from Flat Supervision RuTermEval (Track 3) lemm. inc. + early dmg Scoreboard Class-agnostic F1 0.5875 #1 of 1 Archive leaderboard report
Nested Term Recognition from Flat Supervision RuTermEval (Track 3) lemm. inc. + early dmg Scoreboard Weighted F1 0.4547 #1 of 1 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

Contrastive Learning

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