Update README.md

README.md

@@ -127,4 +127,92 @@ for text in examples:
     print(f"Text: {text}")
     print(f"Sentiment: {sentiment} (Score: {score:.3f})")
 
+```
+
+
+To get the sentiment of a piece of text, use the function below. It gets the sentiment of each sentence and then returns a length-weighted average to get the final sentiment
+
+```python
+# Install and import NLTK for sentence tokenization
+import nltk
+nltk.download('punkt', quiet=True)
+from nltk.tokenize import sent_tokenize
+
+def predict_paragraph_sentiment(text):
+    """
+    Predict sentiment for a paragraph by analyzing individual sentences
+    and calculating a weighted average based on sentence length.
+    
+    Args:
+        text (str): The paragraph text to analyze
+        
+    Returns:
+        tuple: (weighted_score, overall_sentiment, sentence_details)
+    """
+    # Break text into sentences using NLTK
+    sentences = sent_tokenize(text)
+    
+    if not sentences:
+        return 0.0, 'Neutral', []
+    
+    # Analyze each sentence
+    sentence_results = []
+    total_chars = len(text)
+    
+    for sentence in sentences:
+        sentence = sentence.strip()
+        if sentence:  # Skip empty sentences
+            score, sentiment = predict_sentence_sentiment(sentence)
+            char_weight = len(sentence) / total_chars
+            
+            sentence_results.append({
+                'sentence': sentence,
+                'score': score,
+                'sentiment': sentiment,
+                'length': len(sentence),
+                'weight': char_weight
+            })
+    
+    # Calculate weighted average
+    if not sentence_results:
+        return 0.0, 'Neutral', []
+    
+    weighted_score = sum(result['score'] * result['weight'] for result in sentence_results)
+    
+    # Determine overall sentiment
+    if weighted_score <= -0.33:
+        overall_sentiment = 'Negative'
+    elif weighted_score >= 0.33:
+        overall_sentiment = 'Positive'
+    else:
+        overall_sentiment = 'Neutral'
+    
+    return weighted_score, overall_sentiment, sentence_results
+
+# Test the paragraph sentiment function
+test_paragraph = """
+The IT support experience was mixed today. The initial wait time was absolutely terrible - I was on hold for over an hour! 
+However, once I got through to Mike, he was fantastic. He quickly diagnosed the issue with my VPN connection and walked me 
+through the solution step by step. The whole resolution took about 15 minutes once we started working on it. 
+While the wait was frustrating, I'm satisfied with the technical support I received.
+"""
+
+print("=== Paragraph Sentiment Analysis Example ===")
+print(f"Text: {test_paragraph.strip()}")
+print("\n" + "="*80)
+
+weighted_score, overall_sentiment, sentence_details = predict_paragraph_sentiment(test_paragraph)
+
+print(f"\nOVERALL RESULTS:")
+print(f"Weighted Score: {weighted_score:.3f}")
+print(f"Overall Sentiment: {overall_sentiment}")
+
+print(f"\nSENTENCE BREAKDOWN:")
+for i, detail in enumerate(sentence_details, 1):
+    print(f"{i}. \"{detail['sentence']}\"")
+    print(f"   Score: {detail['score']:.3f} | Sentiment: {detail['sentiment']} | Weight: {detail['weight']:.3f}")
+    print()
+
+print("="*80)
+
 ```