SemEval-2019 Task 2: Unsupervised Lexical Frame Induction
The task focuses on automatic discovery of semantic frames — groups of verbs and their argument structures that describe similar situations — without supervision.
It’s inspired by FrameNet and VerbNet, but participants must induce frames directly from raw linguistic data (syntactic and morphological information only absent of semantic annotations).
Behrang QasemiZadeh, Miriam R. L. Petruck, Regina Stodden, Laura Kallmeyer, and Marie Candito. SemEval-2019 Task 2: Unsupervised Lexical Frame Induction.
In Proceedings of the 13th International Workshop on Semantic Evaluation (SemEval-2019), pages 16–30. ACL Anthology | DOI: 10.18653/v1/S19-2003
In Proceedings of the 13th International Workshop on Semantic Evaluation (SemEval-2019), pages 16–30. ACL Anthology | DOI: 10.18653/v1/S19-2003
The task resources:
Selected sentences from WSJ Treebank:
- Sentences augmented with dependency parses and morphological annotations.
- Gold standards: FrameNet 1.7 and VerbNet 3.2 annotations (for evaluation only).
- Evaluation metric: Clustering metrics comparing system outputs to gold frames/roles (e.g., B-Cubed F-score).
(Data can be obtained from LDC)
- + The CodaLab page of the task is available at SemEval 2019 task 2 on Unsuperivsed Lexical Frame Induction
- + The scorer for the task is available for download from http://data-science-expert.de/lr/semeval2019-task2/semeval-2019-task2-scorer.zip
- + The public trial data is available from http://data-science-expert.de/lr/semeval2019-task2/trial-public.zip
Explore truncated data in KWIC view, e.g.:
- - Instances of gold frame annotations;
- - List of annotated frame categories and their frequency;
- - Use the CQL query "[goldl=".*" & syntax="RootInDisc"] within <frame type="Change-position-on-a-scale" />" to get all the verb forms that are annotated as frame type change-position-on-a-scale.
This page last edited on 15 October 2025.
