Create README.md

README.md

@@ -0,0 +1,84 @@
+---
+license: mit
+language:
+- ps
+pretty_name: Pashto Emotion Analysis - Cleaned Dataset
+size_categories: n<1K
+task_categories:
+- text-classification
+---
+
+# Pashto Emotion Dataset (Cleaned Format)
+
+This dataset contains **897 Pashto sentences** labeled for emotion analysis. It is part of a larger project to create resources for Pashto natural language processing.
+
+The data has been cleaned and formatted for easy use in various training and analysis pipelines.
+
+## Dataset Structure
+
+Each entry in the dataset is a JSON object with the following fields:
+
+| Field | Type | Description |
+| :--- | :--- | :--- |
+| `id` | `string` | A unique identifier for the example (e.g., `ex_000001`). |
+| `text` | `string` | The original Pashto sentence. |
+| `emojis` | `list[string]` | A list of emojis present in the text. |
+| `emotion` | `string` | The classified emotion (e.g., `مثبت`, `منفي`, `خنثی`). |
+| `intensity` | `float` | The intensity of the emotion, on a scale from `0.0` to `1.0`. |
+| `cot` | `string` | The "Chain of Thought" reasoning from the DeepSeek model explaining the classification. |
+
+### Example Record
+```json
+{
+  "id": "ex_000001",
+  "text": "او احتمال لري چې دا به غوره وي 🙂",
+  "emojis": ["🙂"],
+  "emotion": "مثبت",
+  "intensity": 0.6,
+  "cot": "احتمال ښیي چې وینا کونکی د یوې مثبتې پایلې هیله لري، مګر ډاډ نلري — 'احتمال لري' د ناڅرګندتیا ښودنه کوي، خو 🙂 دا ناڅرګندتیا نرمه کوي."
+}
+```
+
+## Usage
+
+Load the dataset directly from Hugging Face Datasets:
+
+```python
+from datasets import load_dataset
+
+# Load the dataset
+dataset = load_dataset("nassimjp/pashto-emotion-cleaned", split="train")
+
+# Access an example
+print(dataset[0]["text"])
+print(f"Emotion: {dataset[0]['emotion']}, Intensity: {dataset[0]['intensity']}")
+print(f"CoT: {dataset[0]['cot']}")
+```
+
+## Statistics
+
+- **Total Examples:** 897
+- **Average CoT Length:** 120 characters
+- **Examples with Emojis:** 897 (100%)
+
+### Emotion Distribution
+
+| Emotion | Count | Percentage |
+| :--- | :--- | :--- |
+| خنثی (Neutral) | 347 | 38.7% |
+| مثبت (Positive) | 340 | 37.9% |
+| منفي (Negative) | 116 | 12.9% |
+| طنز (Playful/Sarcastic) | 75 | 8.4% |
+| مخلوط (Mixed) | 15 | 1.7% |
+| حیرانتیا (Surprise) | 3 | 0.3% |
+
+## Related Datasets
+
+This "cleaned" version is one of three formats available for this data. Depending on your use case, you might prefer one of the other formats:
+
+- **For standard fine-tuning (prompt/response):** [nassimjp/pashto-emotion-train](https://huggingface.co/datasets/nassimjp/pashto-emotion-train)
+- **For chat model fine-tuning (conversations list):** [nassimjp/pashto-emotion-chat](https://huggingface.co/datasets/nassimjp/pashto-emotion-chat)
+
+## License
+
+This dataset is released under the MIT License.