use Perceptron classifier which supports partial_fit

core/data_provider/twitter_provider.py

@@ -27,7 +27,7 @@ def _get_api_status_updates(self, user_id):
 
     def _get_dataset_status_updates(self, dataset_path):
         status_updates = []
-        with open(dataset_path, 'r') as dataset_file:
+        with open(dataset_path, 'r', encoding='utf8') as dataset_file:
             csv_reader = csv.DictReader(dataset_file)
             for row in csv_reader:
                 status_updates.append(StatusUpdate.from_dict(row))

core/training/decision_tree.py

@@ -1,8 +1,8 @@
-from sklearn import tree
+from sklearn import linear_model
 
 
 def train_classifier(samples, labels):
-    classifier = tree.DecisionTreeClassifier()
-    classifier = classifier.fit(samples, labels)
+    classifier = linear_model.Perceptron()
+    classifier = classifier.partial_fit(samples, labels, classes=labels)
 
     return classifier