Checklist IOAI Indonesia 2025 National Training Camp 2 – Simulations · 2.3 task
Tweet Emotion Class and Intensity
Predict the emotion class and a continuous intensity score for each tweet.
The task
Given a tweet, predict its emotion class (fear, anger, joy or sadness) and an emotion intensity between 0 and 1, where higher values indicate stronger expression.
Abridged by SOTA from the official materials. The official statement has the exact rules, and it wins wherever this summary differs.
At a glance
- You get
train.csvandtest.csvwith columns tweet, class, intensity (labels only in train).- You submit
- CSV with columns ID, class, intensity.
- Scoring
- Average of the F1 score on class and the R² score on intensity.
- Rules
- Transformer-based models are not allowed; sequence modelling must use LSTM, GRU or a vanilla RNN.
- Pre-trained word2vec or fastText embeddings are allowed.
- Format
- Pelatnas 2 (2025) Simulation 2.3 on Kaggle, 12 July 2025, 06:00–11:00 UTC.