# analyze_sentiment.py # This script analyzes the sentiment of the summarized content using the Hugging Face Transformers library. from transformers import pipeline # Load zero-shot classification pipeline classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli") def analyze_summary(summary): """ Analyze the sentiment of the given summary using zero-shot classification. Returns a tuple of (sentiment, score). """ try: if not summary.strip(): return "No input provided.", 0.0 candidate_labels = ["positive", "neutral", "negative"] result = classifier(summary, candidate_labels) sentiment = result['labels'][0].capitalize() score = float(result['scores'][0]) return sentiment, score except Exception as e: return f"Error analyzing sentiment: {str(e)}", 0.0