SWOT Analysis Model based on DistilBERT

This repository hosts a fine-tuned version of distilbert-base-uncased, specifically trained to classify SWOT elements (Strength, Weakness, Opportunity, Threat) in Amazon product reviews of smartphones. This model serves as a "Synthetic Expert", with annotations derived from a combination of GPT-4 generated labels and human labeling.

Model Training and Data

Base Model: distilbert-base-uncased
Dataset: 9,545 Amazon product reviews.
- Annotations:
  - GPT-4 generated labels for 9,045 reviews.
  - Human-labeled data for 500 reviews as a baseline.
Task: Multi-label classification of SWOT elements.

How to Use

This model can be directly loaded via the Hugging Face Transformers library:

from transformers import AutoModelForSequenceClassification, AutoTokenizer

# Import model and tokenizer
model = AutoModelForSequenceClassification.from_pretrained('jcaponigro/SWOT_Classifier')
tokenizer = AutoTokenizer.from_pretrained('jcaponigro/SWOT_Classifier')

# Example of model usage
text = "Your text for SWOT analysis."
encoded_input = tokenizer(text, return_tensors='pt')
output = model(**encoded_input)