Source code of competition in EmotionGIF 2020 (ACL 2020 Workshop). We won the third place and report can be referred here.

Given the labeled training data, you will need to recommend GIF categories for all the tweets in the unlabeled evaluation dataset. For each tweet, you need to recommend exactly 6 categories.

Our dataset is collected from Twitter, and includes 40K samples. We provide three files: training data, development data that can be used during practice, and evaluation data used for ranking the submissions.

The metric that will be used to evaluate entries is Mean Recall at k, with k=6 (MR@6).