A data scientist is building a bigram model for a medical text corpus. The vocabulary size is 50,000 words. After training, the model assigns zero probability to many bigrams. To ensure the model never assigns zero probability, which technique should be applied?

Use maximum likelihood estimation

Remove rare words from the vocabulary

In an n-gram model, what does the 'n' represent?

The number of sentences in the training data

The number of words in the context used to predict the next word

The number of times a word appears in the training corpus

The total number of unique words in the vocabulary

What is the primary goal of a language model in natural language processing?

To translate text from one language to another

To classify text into predefined categories

To convert text into a vector representation

To assign a probability to a sequence of words

A team is comparing two trigram language models on a held-out test set. Model A has a perplexity of 120, and Model B has a perplexity of 95. Which conclusion is most appropriate?

The perplexity values are too close to draw any conclusion

Model B is better because lower perplexity indicates better predictive performance

Model A is better because higher perplexity indicates more confidence

Perplexity cannot be used to compare models of the same order

In speech recognition, how are language models typically used?

To rank candidate word sequences and select the most probable one

To remove background noise from audio

To convert audio signals into spectrograms

To identify the speaker's gender

Dell Data Science Optimize

Domain 3Objective 3

Language Modeling 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.

26questions here

6free pages

6concepts

20%of the exam

Questions 21–25

21application · medium
A data scientist is building a bigram model for a medical text corpus. The vocabulary size is 50,000 words. After training, the model assigns zero probability to many bigrams. To ensure the model never assigns zero probability, which technique should be applied?
Select an answer first
22foundation · easy
In an n-gram model, what does the 'n' represent?
Select an answer first
23foundation · easy
What is the primary goal of a language model in natural language processing?
Select an answer first
24application · medium
A team is comparing two trigram language models on a held-out test set. Model A has a perplexity of 120, and Model B has a perplexity of 95. Which conclusion is most appropriate?
Select an answer first
25foundation · easy
In speech recognition, how are language models typically used?
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.