Shared tasks in natural language processing
Course organizers: Francis Tyers, Mark Fishel
Code: MTAT.06.046
Load: 6 credit points
The Special Interest Group of the Association of Computational Linguistics (ACL) on Computational Natural Language Learning (SIGNLL) organises a yearly shared task on topics in natural language learning.
A shared task is a competition where training and test data are provided by the organisers and participating systems compete to get the best results. Participating systems are evaluated and compared in a systematic way and the results are presented at an international conference and in the proceedings.
These shared tasks have been run since 1999 and are a great opportunity to take part in an international competition.
In 2017 there are two shared tasks:
- CoNLL-SIGMORPHON-2017 Shared Task: Universal Morphological Reinflection
- CoNLL-2017 Shared Task: Multilingual Parsing from Raw Text to Universal Dependencies
The idea of this course is to prepare you for participating in one of these two shared tasks, and gain great glory and fame for TÜ!