#SocialNLP EmotionGIF 2020 Challenge
The EmotionGIF 2020 Competition has ended.
The big winner, out of a total of 39 participating teams, is Team Mojitok of Platfarm Inc., South Korea.
They have achieved MR@6 scores of 0.6247 (Round 1) and 0.6255 (Round 2). The leaderboards can be viewed here.
Congratulations to the winning team and to all other participants!
EmotionGIF 2020
Welcome to EmotionGIF 2020, the shared task of SocialNLP 2020 (in conjunction with ACL 2020).
With the proliferation of animated GIFs in social networks, their importance as a way to express affective states, including emotions, is growing. Given their expressiveness, GIFs offer a more nuanced and precise way to convey emotions.
This year we present a novel challenge. Our dataset contain 40,000 tweets and their GIF responses, labeled with the categories of the animated GIFs. The challenge: given unlabeled tweets, predict the category(ies) of a GIF response.
Fill the form to participate.
Important: Technical Papers for this task will not be archived
News
June 9: check the workshop session program.
June 5: the leaderboard for Round 2 has been published.
June 11: the leaderboard for Round 1 has been published!
June 8: Due to this year's unique circumstances, we have opened another round ("Round 2") of the competition. See dates and FAQ.
May 15: a ZIP file with all the GIF animations use in the training data can be downloaded here.
May 6, 2020: The challenge is launched. See important dates. Fill the form to participate.
Follow us on Twitter for updates!
Competition Tracks
Main Track: this is the focus of the competition; participants are not allowed to use external data or custom-trained models in their solutions. Participants are allowed to use publicly available pretrained models and embeddings, such as BERT and GloVe.
Auxiliary Track: no limit on use of external data or custom-trained models.
Participants must participate in the Main Track. Submission to the auxiliary track is optional.
There will be separate leaderboards for each track.