Update README.md

README.md

@@ -93,11 +93,11 @@ The output will be a list of recognized entities with their entity type, score,
 ]
 ```
 
-In some cases, we are getting multiple same entity groups so to join please use below code:
+In some cases, we are getting multiple same entity groups, so to join, please use below code:
 
 ```python
 
-def merge_consecutive_entities(entities):
+def merge_consecutive_entities(entities, text):
     entities = sorted(entities, key=lambda x: x['start'])
     merged_entities = []
     current_entity = None
@@ -107,12 +107,11 @@ def merge_consecutive_entities(entities):
             current_entity = entity
         elif (
             entity['entity_group'] == current_entity['entity_group'] and
-            (entity['start'] <= current_entity['end']) 
+            (entity['start'] <= current_entity['end'])
         ):
-            new_word = entity['word']
-            if not current_entity['word'].endswith(new_word):
-                current_entity['word'] += " " + new_word
+            # Merge based on start and end positions in the text
             current_entity['end'] = max(current_entity['end'], entity['end'])
+            current_entity['word'] = text[current_entity['start']:current_entity['end']]
             current_entity['score'] = (current_entity['score'] + entity['score']) / 2  
         else:
             merged_entities.append(current_entity)
@@ -123,6 +122,7 @@ def merge_consecutive_entities(entities):
     return merged_entities
 
 
+
 from transformers import pipeline
 
 # Load the model
@@ -140,7 +140,7 @@ text = ("A 48-year-old female presented with vaginal bleeding and abnormal Pap s
         "hysterectomy with salpingo-oophorectomy which demonstrated positive spread to the pelvic "
         "lymph nodes and the parametrium.")
 result = pipe(text)
-final_result=merge_consecutive_entities(result)
+final_result=merge_consecutive_entities(result,text)
 print(final_result)
 
 ```