debug

generic_ner.py

@@ -206,12 +206,11 @@ def add_spaces_around_punctuation(text):
     return re.sub(r"([{}])".format(re.escape(all_punctuation)), r" \1 ", text)
 
 
-# Function to find the closest entity to the "comp.function"
 def attach_comp_to_closest(entities):
     # Define valid entity types that can receive a "comp.function" or "comp.name" attachment
     valid_entity_types = {"org", "pers", "org.ent", "pers.ind"}
 
-    # Separate "comp.function" entities from other entities
+    # Separate "comp.function" and "comp.name" entities from other entities
     comp_entities = [ent for ent in entities if ent["entity"].startswith("comp")]
     other_entities = [ent for ent in entities if not ent["entity"].startswith("comp")]
 
@@ -221,20 +220,27 @@ def attach_comp_to_closest(entities):
 
         # Find the closest non-"comp.function" entity that is valid for attaching
         for other_entity in other_entities:
-            distance = abs(comp_entity["start"] - other_entity["end"])  # Calculate the distance
+            distance = abs(
+                comp_entity["start"] - other_entity["end"]
+            )  # Calculate the distance
 
             # Ensure that the other entity's type is valid for the attachment
-            if distance < min_distance and other_entity["entity"].split(".")[0] in valid_entity_types:
+            if (
+                distance < min_distance
+                and other_entity["entity"].split(".")[0] in valid_entity_types
+            ):
                 # Additional context check: Ensure the comp_entity's text doesn't conflict with the target entity's text
                 if not conflicting_context(comp_entity, other_entity):
                     min_distance = distance
                     closest_entity = other_entity
 
-        # Attach the "comp.function" or "comp.name" entity using the suffix of 'entity' field if a valid entity is found
+        # Attach the "comp.function" entity using the suffix of 'entity' field if a valid entity is found
         if closest_entity:
             # Extract the suffix (e.g., "comp.title" becomes "title")
             suffix = comp_entity["entity"].split(".")[-1]
-            closest_entity[suffix] = comp_entity["text"]  # Attach the text using the suffix as the key
+            closest_entity[suffix] = comp_entity[
+                "text"
+            ]  # Attach the text using the suffix as the key
 
     return other_entities
 
@@ -247,15 +253,25 @@ def conflicting_context(comp_entity, target_entity):
     comp_text = comp_entity["text"].lower()
     target_text = target_entity["text"].lower()
 
-    # Simple conflict check: don't attach a name if the target entity text contains a different person or entity name
-    # Example: If target entity is 'Marie Antoinette', don't attach 'Maximilien Robespierre'
-    if "pers" in target_entity["entity"] and comp_text not in target_text:
-        return True  # Conflict: Different person/entity
+    # Case 1: Avoid attaching the name "Washington" to "George Washington" if it relates to a location.
+    if "pers" in target_entity["entity"] and comp_entity["entity"].startswith(
+        "comp.name"
+    ):
+        # Ensure we allow attaching "Washington" to "George Washington" but not to other irrelevant contexts
+        if comp_text not in target_text and target_text.startswith("george washington"):
+            return True  # Conflict: Unrelated or partial name mismatch
+
+    # Case 2: Ensure the target entity text matches the role or function (e.g., "President of the United States").
+    if "pers" in target_entity["entity"] and comp_entity["entity"].startswith(
+        "comp.function"
+    ):
+        # Ensure the function context makes sense (e.g., attach 'President' to 'George Washington')
+        if "president" not in comp_text and "george washington" in target_text:
+            return True  # Conflict: Unrelated function
 
     return False  # No conflict
 
 
-
 def postprocess_entities(entities):
     # Step 1: Filter entities with the same text, keeping the one with the most dots in the 'entity' field
     entity_map = {}