What NLP actually does - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding NLP Tasks

Which of the following best describes a primary task of Natural Language Processing (NLP)?

A Converting images into text descriptions

B Translating human language into machine-readable data for understanding and response

C Optimizing hardware performance for faster computing

D Storing large amounts of data efficiently

💡 Hint

Think about what NLP does with human language.

NLP focuses on making machines understand and work with human language by converting it into a form they can process.

❓ Predict Output

intermediate

Output of Tokenization in NLP

What is the output of this Python code using NLTK for tokenization?

NLP

import nltk
nltk.download('punkt')
from nltk.tokenize import word_tokenize
text = "Hello world! NLP is fun."
tokens = word_tokenize(text)
print(tokens)

A ['Hello world', 'NLP is fun']

B ['Hello world!', 'NLP is fun.']

C ['Hello', 'world', '!', 'NLP', 'is', 'fun', '.']

D ['Hello', 'world!', 'NLP', 'is', 'fun.']

💡 Hint

Tokenization splits text into words and punctuation separately.

The word_tokenize function splits the sentence into words and punctuation marks as separate tokens.

❓ Model Choice

advanced

Choosing the Right NLP Model for Sentiment Analysis

Which model is best suited for analyzing the sentiment (positive or negative feeling) of movie reviews?

A A convolutional neural network (CNN) trained on labeled text data

B A k-means clustering model for grouping text

C A linear regression model predicting numerical values

D A decision tree model for image classification

💡 Hint

Sentiment analysis needs a model that understands text and classifies feelings.

CNNs can learn patterns in text sequences and are effective for sentiment classification tasks.

❓ Metrics

advanced

Evaluating NLP Classification Accuracy

You trained an NLP model to classify emails as spam or not spam. Which metric best tells you how many emails were correctly classified overall?

A Accuracy

B Recall

C Precision

D F1 Score

💡 Hint

Think about the metric that measures correct predictions out of all predictions.

Accuracy measures the proportion of total correct predictions (both spam and not spam) out of all emails.

🔧 Debug

expert

Debugging a Named Entity Recognition (NER) Model Output

Given this Python code snippet using spaCy for NER, what error or issue will occur when running it?

NLP

import spacy
nlp = spacy.load('en_core_web_sm')
doc = nlp('Apple is looking at buying U.K. startup for $1 billion')
for ent in doc.ents:
    print(ent.text, ent.label_)

# Later code
print(doc.entities)

A TypeError because 'ent.label_' is not a string

B Prints all entities correctly without error

C SyntaxError due to missing colon in for loop

D AttributeError because 'Doc' object has no attribute 'entities'

💡 Hint

Check spaCy's documentation for the correct attribute name to access entities.

The correct attribute is 'ents', not 'entities'. Using 'doc.entities' causes an AttributeError.