|
|
--- |
|
|
task_categories: |
|
|
- text-classification |
|
|
--- |
|
|
|
|
|
# INTENT-ES Dataset |
|
|
|
|
|
This dataset contains Spanish language texts labeled with: |
|
|
|
|
|
1. Intent classification (13 classes): |
|
|
- 0: Threat |
|
|
- 1: Criticism |
|
|
- 2: Desire |
|
|
- 3: Praise |
|
|
- 4: Emotional |
|
|
- 5: Informative |
|
|
- 6: Command |
|
|
- 7: Personal Opinion |
|
|
- 8: Request |
|
|
- 9: Question |
|
|
- 10: Promise |
|
|
- 11: Sarcasm/Joke |
|
|
- 12: Suggestion |
|
|
|
|
|
|
|
|
2. Sentiment analysis (3 classes): |
|
|
- 0: NEG |
|
|
- 1: NEU |
|
|
- 2: POS |
|
|
|
|
|
|
|
|
3. Emotion detection (7 classes): |
|
|
- 0: others |
|
|
- 1: joy |
|
|
- 2: sadness |
|
|
- 3: anger |
|
|
- 4: surprise |
|
|
- 5: disgust |
|
|
- 6: fear |
|
|
|
|
|
|
|
|
Each entry in the dataset includes the original text, the intent label and its text description, |
|
|
the sentiment and emotion labels (both as numeric IDs and as text), and confidence scores for sentiment and emotion classification. |
|
|
|
|
|
## Label Distributions |
|
|
|
|
|
The dataset has 6493 training examples and 1620 test examples. |
