Aspect-based sentiment analysis in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load the aspect-based sentiment analysis model.

NLP

from transformers import pipeline

nlp = pipeline([1], model='nlptown/bert-base-multilingual-uncased-sentiment')

Drag options to blanks, or click blank then click option'

A'aspect-based-sentiment-analysis'

B'sentiment-analysis'

C'text-classification'

D'ner'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'sentiment-analysis' instead of 'aspect-based-sentiment-analysis'.

Using 'ner' which is for named entity recognition.

The pipeline type for aspect-based sentiment analysis is 'aspect-based-sentiment-analysis'.

2fill in blank

medium

Complete the code to extract aspect terms from the input text.

NLP

text = "The battery life of this phone is amazing but the screen is dull."
results = nlp(text)
aspects = [item['[1]'] for item in results]

Drag options to blanks, or click blank then click option'

Alabel

Baspect

Cscore

Dsentiment

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'label' which usually indicates sentiment polarity.

Using 'score' which is the confidence value.

The key 'aspect' holds the aspect term extracted from the text.

3fill in blank

hard

Fix the error in the code to correctly get the sentiment label for each aspect.

NLP

for item in results:
    print(item['[1]'])

Drag options to blanks, or click blank then click option'

Alabel

Bsentiment

Caspect

Dscore

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'aspect' which is the feature, not the sentiment.

Using 'sentiment' which is not a key in the result dictionary.

The sentiment label is stored under the key 'label' in each result item.

4fill in blank

hard

Fill both blanks to create a dictionary mapping each aspect to its sentiment label.

NLP

aspect_sentiments = {item[[1]]: item[[2]] for item in results}

Drag options to blanks, or click blank then click option'

A'aspect'

B'label'

C'sentiment'

D'score'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'sentiment' which is not a valid key.

Using 'score' which is a confidence value, not a label.

The dictionary keys are aspect terms ('aspect') and values are sentiment labels ('label').

5fill in blank

hard

Fill all three blanks to filter aspects with positive sentiment and create a list of their terms.

NLP

positive_aspects = [item[[1]] for item in results if item[[2]] == [3]]

Drag options to blanks, or click blank then click option'

A'aspect'

B'label'

C'positive'

D'score'

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'score' instead of 'label' for sentiment filtering.

Using 'sentiment' which is not a valid key.

We select the 'aspect' terms where the 'label' equals 'positive'.