NER and logging

.gitignore

@@ -1,2 +1,3 @@
 # EddyGiusepe
 venv_logging/
+reconhecimento_de_entidade.log

4_Entity_and_logging.py

@@ -0,0 +1,84 @@
+#!/usr/bin/env python3
+
+"""
+Data Scientist.: Dr.Eddy Giusepe Chirinos Isidro
+
+Objetivo: Neste script utilizamos um modelo pré-treinado para extrair
+          Entidades e usamos o pacote logging do python para registrar 
+          nossos LOGs. 
+"""
+import logging
+from transformers import pipeline
+
+class EntityRecognizer:
+    def __init__(self, model_name="Babelscape/wikineural-multilingual-ner"):  # https://huggingface.co/Babelscape/wikineural-multilingual-ner
+        self.model = self.load_model(model_name)
+        self.logger = self.setup_logger()
+
+    def load_model(self, model_name="Babelscape/wikineural-multilingual-ner"):
+        # Carrego o modelo pré-treinado do Hugging Face:
+        return pipeline("ner", model=model_name, tokenizer=model_name)
+
+    def setup_logger(self):
+        # Configuração de Logs:
+        logger = logging.getLogger(__name__)
+        logger.setLevel(logging.INFO)
+
+        formatter = logging.Formatter('%(asctime)s - %(levelname)s - %(message)s')
+
+        file_handler = logging.FileHandler('reconhecimento_de_entidade.log')
+        file_handler.setLevel(logging.INFO)
+        file_handler.setFormatter(formatter)
+
+        logger.addHandler(file_handler)
+
+        return logger
+
+    def recognize_entities(self, text):
+        # Use o modelo NER pré-treinado para reconhecer entidades no texto:
+        entities = self.model(text)
+
+        recognized_entities = []
+
+        for entity in entities:
+            entity_text = entity['word']
+            entity_type = entity['entity']
+            recognized_entities.append((entity_text, entity_type))
+
+        self.logger.info(f"Entidades reconhecidas: {recognized_entities}")
+
+        return recognized_entities
+
+    def process_classification_result(self, tokens_and_tags):
+        result = {}
+        current_type = None
+        current_entity = ""
+
+        for token, tag in tokens_and_tags:
+            if tag.startswith("B-"):
+                if current_type is not None and current_entity:
+                    result[current_entity] = current_type
+                current_type = tag[2:]
+                current_entity = token
+            elif tag.startswith("I-"):
+                current_entity += " " + token
+
+        if current_type is not None and current_entity:
+            result[current_entity] = current_type
+
+        return result
+
+
+if __name__ == "__main__":
+    # Exemplo de uso:
+    #model_name = "Babelscape/wikineural-multilingual-ner"
+    #text = "O Eddwin e a Karina foram para Estados Unidos a estudar em Harvard."
+    text = "Eddy e Karina compraram uns tênis na loja Nike."
+    entity_recognizer = EntityRecognizer() # entity_recognizer = EntityRecognizer(model_name)
+    recognized = entity_recognizer.recognize_entities(text)
+    print(recognized)
+    print("🤗🤗🤗")
+
+    result = entity_recognizer.process_classification_result(recognized)
+    result = {k.replace(" ##", ""): v for k, v in result.items()}  # Remove '##' from keys
+    print(result)