Choosing number of topics in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the library needed for topic modeling.

NLP

from sklearn.decomposition import [1]

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

APCA

BKMeans

CNMF

DTSNE

Attempts:

3 left

💡 Hint

Common Mistakes

Choosing PCA or KMeans which are not used for topic modeling here.

The NMF class from sklearn.decomposition is used for topic modeling by factorizing the document-term matrix.

2fill in blank

medium

Complete the code to create an NMF model with a specified number of topics.

NLP

model = NMF(n_components=[1], random_state=42)

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

A5

B0.5

C'five'

D-1

Attempts:

3 left

💡 Hint

Common Mistakes

Using a float or string instead of an integer for number of topics.

The n_components parameter expects an integer number of topics, such as 5.

3fill in blank

hard

Fix the error in the code to fit the NMF model on the document-term matrix X.

NLP

W = model.[1](X)

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

Afit_predict

Bfit_transform

Ctransform

Dpredict

Attempts:

3 left

💡 Hint

Common Mistakes

Using transform without fitting first causes an error.

The fit_transform method fits the model and returns the topic distribution matrix W.

4fill in blank

hard

Fill both blanks to compute reconstruction error and print it.

NLP

error = model.[1]
print('Reconstruction error:', model.[2])

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

Areconstruction_err_

Breconstruction_error_

Creconstruction_err

Dreconstruction_error

Attempts:

3 left

💡 Hint

Common Mistakes

Using incorrect attribute names like reconstruction_error or missing underscore.

The attribute reconstruction_err_ stores the reconstruction error after fitting the model.

5fill in blank

hard

Fill all three blanks to create a dictionary of topic words with their weights for the first topic.

NLP

topic_words = {word: model.components_[0, [1]] for [2], word in enumerate(vectorizer.get_feature_names_out()) if [3] < 10}

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

Ai

Bidx

Dindex

Attempts:

3 left

💡 Hint

Common Mistakes

Mixing variable names or using wrong variable names for indices.

We use idx as the index to access weights, idx as the index variable in enumeration, and filter words with index less than 10.