def test_label(nlp: BertPipeline):
# We add the pattern recognizer to the pipeline.
pattern_recognizer = AttentionPatternRecognizer()
nlp.pattern_recognizer = pattern_recognizer
text_1 = ("We are great fans of Slack, but we wish the subscriptions "
"were more accessible to small startups.")
text_2 = "The Slack often has bugs."
text_3 = "best of all is the warm vibe"
aspect = "slack"
examples = [
Example(text_1, aspect),
Example(text_2, aspect),
Example(text_3, aspect)
]
tokenized_examples = nlp.tokenize(examples)
input_batch = nlp.encode(tokenized_examples)
output_batch = nlp.predict(input_batch)
labeled_examples = nlp.label(tokenized_examples, output_batch)
labeled_examples = list(labeled_examples)
labeled_1, labeled_2, labeled_3 = labeled_examples
assert labeled_1.sentiment == Sentiment.positive
assert labeled_2.sentiment == Sentiment.negative
assert isinstance(labeled_1.scores, list)
assert np.argmax(labeled_1.aspect_representation.look_at) == 5
assert np.argmax(labeled_2.aspect_representation.look_at) == 1
# We need to calibrate the model. The prediction should be neutral.
# In fact, the model does not recognize the aspect correctly.
assert labeled_3.sentiment == Sentiment.positive
assert np.allclose(labeled_3.aspect_representation.look_at,
[1.0, 0.16, 0.50, 0.54, 0.34, 0.39, 0.12],
atol=0.01)
def test_get_completed_task(nlp: BertPipeline):
text = ("We are great fans of Slack.\n"
"The Slack often has bugs.\n"
"best of all is the warm vibe")
# Make sure we have defined a text_splitter, even naive.
nlp.text_splitter = lambda text: text.split('\n')
task = nlp.preprocess(text, aspects=['slack', 'price'])
tokenized_examples = task.batch
input_batch = nlp.encode(tokenized_examples)
output_batch = nlp.predict(input_batch)
aspect_span_labeled = nlp.label(tokenized_examples, output_batch)
completed_task = nlp.get_completed_task(task, aspect_span_labeled)
assert len(completed_task.batch) == 6
assert completed_task.indices == [(0, 3), (3, 6)]
slack, price = completed_task
assert slack.text == price.text == text
# The sentiment among fragments are different. We normalize scores.
assert np.allclose(slack.scores, [0.06, 0.46, 0.48], atol=0.01)
# Please note once gain that there is a problem
# with the neutral sentiment, model is over-fitted.
assert np.allclose(price.scores, [0.06, 0.42, 0.52], atol=0.01)