Update app.py

app.py

@@ -30,6 +30,13 @@ class CRF_POS_Tagger:
             "ness", "ous", "ship", "y", "es", "s"
         ]
 
+        self.interjections = {
+        'ah', 'alas', 'aha', 'bravo', 'darn', 'drat', 'eh', 'eek', 'eww', 
+        'gosh', 'ha', 'hooray', 'hmm', 'huh', 'oops', 'ouch', 'phew', 
+        'pow', 'yay', 'whoa', 'wow', 'yikes', 'yippee', 'uh', 'um', 
+        'hey', 'hello'
+        }
+
         self.prefix_pattern = f"^({'|'.join(self.prefixes)})"
         self.suffix_pattern = f"({'|'.join(self.suffixes)})$"
 
@@ -42,7 +49,7 @@ class CRF_POS_Tagger:
         self.X_test = self.X[self.split:]
         self.y_test = self.y[self.split:]
         self.crf_model = sklearn_crfsuite.CRF(algorithm='lbfgs', c1=0.1, c2=0.1, max_iterations=100, all_possible_transitions=True)
-        self.train()
+        # self.train()
 
     def word_splitter(self, word):
         prefix = ""
@@ -148,8 +155,9 @@ class CRF_POS_Tagger:
         
         return features
 
-    def train(self):
-        self.crf_model.fit(self.X_train, self.y_train)
+    def train(self, data=None):
+        X_train, y_train = zip(*data) if data else self.X_train, self.y_train
+        self.crf_model.fit(X_train, y_train)
     
     def predict(self, X_test):
         return self.crf_model.predict(X_test)
@@ -161,7 +169,8 @@ class CRF_POS_Tagger:
         self.predicted_tag.extend([item for sublist in y_pred for item in sublist])
         return metrics.flat_accuracy_score(y_test, y_pred)
 
-    def cross_validation(self, data):
+    def cross_validation(self):
+        data = list(zip(self.X, self.y))  
         accuracies = []
         for i in range(5):
             n1 = int(i / 5.0 * len(data))
@@ -218,7 +227,7 @@ class CRF_POS_Tagger:
             
     def tagging(self, input):
         sentence = (re.sub(r'(\S)([.,;:!?])', r'\1 \2', input.strip())).split()
-        sentence_list = [[word] for word in sentence]
+        sentence_list = [[word.lower()] for word in sentence]
         features = [self.word_features(sentence_list, i) for i in range(len(sentence_list))]
 
         predicted_tags = self.crf_model.predict([features])
@@ -226,6 +235,14 @@ class CRF_POS_Tagger:
         return output
 
 tagger = CRF_POS_Tagger()
+
+accuracies, avg_accuracy = tagger.cross_validation()
+print(f"Cross-Validation Accuracies: {accuracies}")
+print(f"Average Accuracy: {avg_accuracy}")
+
+conf_matrix = tagger.con_matrix()
+print(tagger.per_pos_accuracy(conf_matrix))
+
 interface = gr.Interface(fn = tagger.tagging, 
                             inputs = gr.Textbox(
                             label="Input Sentence",