Overview
The UniDive COST Action has announced the second call for participation in the shared tasks on idiomaticity and multiword expressions (MWEs) as part of the PARSEME 2.0 and AdMiRe 2.0 initiatives. These tasks aim to explore the complexities of MWEs, which are phrases whose meanings cannot be directly inferred from their individual words. This opportunity is significant for researchers and practitioners in the AI and ML community, particularly those focused on natural language processing (NLP).
Background & Relevance
Idiomatic expressions and multiword phrases are prevalent in languages and pose unique challenges for computational understanding. MWEs often require contextual interpretation, making them a critical area of study in NLP. The ability to accurately identify and paraphrase these expressions is essential for improving machine understanding of human language. This initiative not only aims to advance research in this area but also fosters collaboration among international teams.
Key Details
- Important Dates:
- 1 October 2025: Release of training data and baseline systems
- 3 December 2025: Publication of test blind data
- 8 December 2025: Submission of system predictions
- 19 December 2025: Evaluation of systems
- 5 January 2026: Deadline for system description papers
- 9-23 January 2026: Reviewing period
- 3 February 2026: Submission deadline for camera-ready papers
- 24-29 March 2026: EACL conference, including the MWE workshop (to be confirmed)
- Location: Morocco (EACL 2026)
- Tracks/Topics:
- Identification and paraphrasing of MWEs
- Multimodal idiomaticity representation
- Links:
- Call for Participation
- Training Data for AdMiRe 2.0
- Trial Data for PARSEME 2.0
- Funding Information
Eligibility & Participation
The shared tasks are open to researchers and teams interested in advancing the understanding of idiomaticity and MWEs. Participants can include those working on multilingual and multimodal technologies, as well as those focusing on specific languages or modalities.
Submission or Application Guidelines
Participants are encouraged to submit their system predictions and description papers according to the provided deadlines. The submission process will involve evaluating systems based on their performance in identifying and paraphrasing MWEs, as well as their ability to interpret idiomatic expressions in a multimodal context.
Additional Context / Real-World Relevance
The study of idiomaticity and MWEs is crucial for enhancing machine understanding of human language, which has applications in various domains, including translation, sentiment analysis, and conversational agents. By engaging in these shared tasks, participants contribute to the broader goal of improving NLP technologies and their applications across different languages and contexts.
Conclusion
Researchers and practitioners in the field of NLP are encouraged to explore this opportunity to participate in the UniDive shared tasks. By contributing to the understanding of idiomaticity and MWEs, participants can help advance the state of the art in language processing technologies. Interested individuals should consider applying and collaborating with others in this exciting area of research.
Category: CFP & Deadlines
Tags: idiomaticity, multiword expressions, natural language processing, EACL, multilingual, machine learning, PARSEME, AdMiRe, text analysis, semantic understanding
