Merge pull request #5200 from RasaHQ/fix-check-test-features

Add check for text features

changelog/5199.bugfix.rst

@@ -0,0 +1 @@
+If no text features are present in ``EmbeddingIntentClassifier`` return the intent ``None``.

data/test/many_intents.md

@@ -2,6 +2,11 @@
 - you are an idiot
 - You lack understanding.
 
+## intent:grett
+- Hello
+- Hi
+- Welcome
+
 ## intent:thank
 - Thanks
 - Thank you

rasa/nlu/classifiers/embedding_intent_classifier.py

@@ -817,6 +817,16 @@ def _calculate_message_sim(
         # transform sim to python list for JSON serializing
         return label_ids, message_sim.tolist()
 
+    @staticmethod
+    def _text_features_present(session_data: SessionDataType) -> bool:
+        return np.array(
+            [
+                f.nnz != 0 if isinstance(f, scipy.sparse.spmatrix) else f.any()
+                for features in session_data["text_features"]
+                for f in features
+            ]
+        ).any()
+
     def predict_label(
         self, message: "Message"
     ) -> Tuple[Dict[Text, Any], List[Dict[Text, Any]]]:
@@ -835,6 +845,12 @@ def predict_label(
 
         # create session data from message and convert it into a batch of 1
         session_data = self._create_session_data([message])
+
+        # if no text-features are present (e.g. incoming message is not in the
+        # vocab), do not predict a random intent
+        if not self._text_features_present(session_data):
+            return label, label_ranking
+
         batch = train_utils.prepare_batch(
             session_data, tuple_sizes=self.batch_tuple_sizes
         )

tests/nlu/classifiers/test_embedding_intent_classifier.py

@@ -278,3 +278,29 @@ async def test_margin_loss_is_not_normalized(
 
     # make sure top ranking is reflected in intent prediction
     assert parse_data.get("intent") == intent_ranking[0]
+
+
+@pytest.mark.parametrize(
+    "session_data, expected",
+    [
+        (
+            {
+                "text_features": [
+                    np.array(
+                        [
+                            np.random.rand(5, 14),
+                            np.random.rand(2, 14),
+                            np.random.rand(3, 14),
+                        ]
+                    )
+                ]
+            },
+            True,
+        ),
+        ({"text_features": [np.array([0, 0, 0])]}, False),
+        ({"text_features": [scipy.sparse.csr_matrix([0, 0, 0])]}, False),
+        ({"text_features": [scipy.sparse.csr_matrix([0, 31, 0])]}, True),
+    ],
+)
+def test_text_features_present(session_data, expected):
+    assert EmbeddingIntentClassifier._text_features_present(session_data) == expected