Shared Task: Cross-Framework Meaning Representation

LTG staff coordinates the Shared Task at the 2019 Conference on Computational Language Learning (CoNLL).

Data-driven parsing into graph-structured representations of sentence meaning is a problem that receives increasing R&D interest in recent years, in part because all things semantic are in fashion, in part because moving beyond rooted trees as the target representation remains a modeling and algorithmic challenge.  To stimulate this development, with a particular emphasis on multi-task learning approaches, LTG staff has teamed up with researchers in The Czech Republic, Denmark, Israel, Sweden, and the US in organizing the Shared Task on Cross-Framework Meaning Representation Parsing (MRP 2019), which has been selected as the ‘system bake-off’ problem by the 2019 Conference on Computational Language Learning (CoNLL), one of the highest-profile competitions in Natural Language Processing.

Published Mar. 16, 2019 1:18 PM - Last modified Aug. 27, 2019 1:29 PM