GloVe embeddings in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load GloVe embeddings from a file.

NLP

glove_vectors = {}
with open('glove.txt', 'r', encoding='utf8') as f:
    for line in f:
        values = line.split()
        word = values[0]
        vector = list(map(float, values[[1]]))
        glove_vectors[word] = vector

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

A3:

B0:

C2:

D1:

Attempts:

3 left

💡 Hint

Common Mistakes

Using 0: which includes the word as a float and causes an error.

Using 2: or 3: which skips some vector values.

The first value in each line is the word, so the vector starts from index 1 onward.

2fill in blank

medium

Complete the code to get the embedding vector for the word 'apple'.

NLP

word = 'apple'
embedding = glove_vectors.get([1], None)

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

A'apple'

B'APPLE'

C'apples'

D'Apple'

Attempts:

3 left

💡 Hint

Common Mistakes

Using uppercase or capitalized versions that do not exist in the dictionary.

Using plural form 'apples' which may not be in the embeddings.

GloVe embeddings are case sensitive and usually lowercase, so 'apple' is correct.

3fill in blank

hard

Fix the error in the code to compute cosine similarity between two GloVe vectors.

NLP

import numpy as np

def cosine_similarity(vec1, vec2):
    dot_product = np.dot(vec1, vec2)
    norm1 = np.linalg.norm(vec1)
    norm2 = np.linalg.norm([1])
    return dot_product / (norm1 * norm2)

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

Avec2

Bvec1 - vec2

Cvec1

Dvec1 + vec2

Attempts:

3 left

💡 Hint

Common Mistakes

Using the first vector twice causing incorrect similarity values.

Using vector addition or subtraction instead of the second vector.

The norm of the second vector should be computed using vec2 to calculate cosine similarity correctly.

4fill in blank

hard

Fill both blanks to create a dictionary of words with vectors longer than 50 dimensions.

NLP

filtered_glove = {word: vec for word, vec in glove_vectors.items() if len(vec) [1] [2]

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

A>

B50

C<

D100

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' which filters vectors shorter than 50.

Using 100 which is not the intended threshold.

We want vectors with length greater than 50, so use '>' and 50.

5fill in blank

hard

Fill all three blanks to create a list of words whose embedding vectors have norm greater than 1.0.

NLP

import numpy as np

words_with_norm = [word for word, vec in glove_vectors.items() if np.linalg.norm(vec) [1] [2]]

result = words_with_norm[:[3]]

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

A>

B1.0

C100

D<

Attempts:

3 left

💡 Hint

Common Mistakes

Using '<' which selects vectors with norm less than 1.0.

Using wrong slice number causing too few or too many words.

We check if norm is greater than 1.0 and then take the first 100 words.