Python Sentiment Analysis with Transformers/ LLM.

[davchuks]

Description

This PR introduces a new dataset for emotional polarity analysis and cleans up redundant files in the repository.

• Added emotion_dataset_v1.csv as the updated canonical dataset for emotion analysis.
• Removed outdated/duplicate CSV files (emotion.csv, emotion_v1.csv) to reduce redundancy and avoid confusion.
• Prepared groundwork for fine-tuning emotional polarity classifiers (Positive/Negative) using transformer models (e.g., Bio_ClinicalBERT).
• Ensured outputs include Emotion_Label mapped to polarity for downstream use.

1. Pull this branch from the repository.
2. Confirm that emotion_dataset_v1.csv exists in AI Guardian/emotion_analysis/.
3. Verify that emotion.csv and emotion_v1.csv are no longer present.
4. Run the notebook Next-Gen_Emotional_Polarity.ipynb to ensure that the new dataset loads correctly and outputs the Emotion_Label column with polarity mapping.

Known Issues

• Evaluation metrics (F1, precision, recall) still need to be run and validated.
• Documentation for mapping emotion → polarity requires final review.