A financial news classifier must distinguish between 'Q3 earnings up 5%' and 'Q3 earnings up five percent'. The model uses a vocabulary of the 10,000 most frequent tokens. How should numbers be handled to maximize classification accuracy?

Convert all numbers to a single placeholder token like ' '.

Keep all numbers as-is in their original form.

Convert numbers to words only when they appear with a percent sign.

Remove all numbers and percentage signs from the text.

Why is lowercasing commonly applied as a text preprocessing step?

To split text into individual tokens

To remove punctuation from the text

To ensure consistency by treating 'Hello' and 'hello' as the same token

To convert words to their root form

A team is building a sentiment model for product reviews. They notice that 'Great' and 'great' are treated as different tokens, increasing vocabulary size. However, they also need to preserve proper nouns like 'Apple' (the company) for entity recognition. What is the best approach?

Keep all text as-is and rely on the model to learn case differences.

Lowercase all text without exception.

Remove all proper nouns from the text.

Lowercase all text except tokens identified as proper nouns by a POS tagger.

Dell Data Science Optimize

Domain 3Objective 2

Text Preprocessing DATA-SCIENCE-OPTIMIZE Practice Questions (Page 5)

Part of the Natural Language Processing (NLP) domain, which accounts for 20% of the DATA-SCIENCE-OPTIMIZE exam.

44questions here

9free pages

15concepts

20%of the exam

Questions 21–25

21application · medium
A financial news classifier must distinguish between 'Q3 earnings up 5%' and 'Q3 earnings up five percent'. The model uses a vocabulary of the 10,000 most frequent tokens. How should numbers be handled to maximize classification accuracy?
Select an answer first
22foundation · easy
Why is lowercasing commonly applied as a text preprocessing step?
Select an answer first
23foundation · easy
Which tool is commonly used for lemmatization in Python's NLTK library?
Select an answer first
24expert · hard
A team is building a sentiment model for product reviews. They notice that 'Great' and 'great' are treated as different tokens, increasing vocabulary size. However, they also need to preserve proper nouns like 'Apple' (the company) for entity recognition. What is the best approach?
Select an answer first
25foundation · easy
Which of the following is an example of a non-alphanumeric symbol that might be removed during preprocessing?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to DATA-SCIENCE-OPTIMIZE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Dell Technologies. “DATA-SCIENCE-OPTIMIZE” is a trademark of its owner, used for identification only.

Text Preprocessing DATA-SCIENCE-OPTIMIZE Practice Questions (Page 5)

Questions 21–25

A financial news classifier must distinguish between 'Q3 earnings up 5%' and 'Q3 earnings up five percent'. The model uses a vocabulary of the 10,000 most frequent tokens. How should numbers be handled to maximize classification accuracy?

Why is lowercasing commonly applied as a text preprocessing step?

Which tool is commonly used for lemmatization in Python's NLTK library?

A team is building a sentiment model for product reviews. They notice that 'Great' and 'great' are treated as different tokens, increasing vocabulary size. However, they also need to preserve proper nouns like 'Apple' (the company) for entity recognition. What is the best approach?

Which of the following is an example of a non-alphanumeric symbol that might be removed during preprocessing?

Keep studying — Natural Language Processing (NLP)