2

Looking for Linguistics Competitions suitable for MSc or PhD students in Computational Linguistics. I don't see anything current on Kaggle, and I am looking into TREC. added: recurring preferred (thank you, @tripleee).

I'm just a student myself, so I would welcome any pointers about where to look that experienced Linguists can suggest.


I thought about whether this is an answerable question and it seems that community resources, like competitions, may have concrete answers (vs "only opinions"). So I'm tentatively raising the question. And I am totally open to being corrected and will un-ask if this won't be helpful to others (or is otherwise ill-suited for the stackexchange Linguistics forum).

  • 1
    SQuAD, the Winograd Schema Challenge, and the WMT workshops for each problem area, eg statmt.org/wmt18/quality-estimation-task.html. – Adam Bittlingmayer Apr 7 '18 at 21:31
  • 1
    Many competitions are one-off, which means answers will inherently be struggling with out of date information. Maybe ask specifically for recurring competitions ...? – tripleee Apr 9 '18 at 7:01
1

There is one that I know of: Low Resource Languages for Emergent Incidents (LORELEI)

Here is the link (also a NIST competition)

| improve this answer | |
  • 1
    Thank you, marked as accepted. I posted an email excerpt from NIST as well. – jgreve Apr 11 '18 at 20:03
1

NIST was kind enough to relay this by email (April 2018):

re Text Retrieval Conference (TREC):

The call for participation in TREC 2018 is still active, so you can sign up to participate this year is you would like. You can learn more about the types of tasks the tracks have focused on in previous years from the Overview papers in the Proceedings. The proceedings are in the Publications section of the TREC web site https://trec.nist.gov/

TREC's focus is on search, and while there is definitely room for linguistic approaches to search, that is not the most frequent approach. Our group at NIST has a sibling project to TREC called the Text Analysis Conference (TAC) that focuses specifically on natural language processing tasks. The TAC link currently points to TAC 2017, but I expect it will be updated for TAC 2018 very soon.

You might also be interested in evaluations run by the NIST Multimodal Information Group.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.