Domain-specific sentiment in NLP - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to load the domain-specific sentiment dataset.

NLP

import pandas as pd
data = pd.read_csv('[1]')

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

Ageneral_sentiment.csv

Bdomain_sentiment.csv

Crandom_data.csv

Duser_reviews.csv

Attempts:

3 left

💡 Hint

Common Mistakes

Loading a general sentiment dataset instead of domain-specific.

Using a file name that does not exist.

The domain-specific sentiment dataset is stored in 'domain_sentiment.csv'. Loading this file ensures the model learns from relevant data.

2fill in blank

medium

Complete the code to split the dataset into features and labels.

NLP

X = data['[1]']
y = data['sentiment']

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

Atext

Blabel

Cscore

Dreview

Attempts:

3 left

💡 Hint

Common Mistakes

Using the label column as features.

Choosing a column unrelated to text.

The feature column containing the text data is named 'text'. This is used as input for the model.

3fill in blank

hard

Fix the error in the code to vectorize the text data using TF-IDF.

NLP

from sklearn.feature_extraction.text import TfidfVectorizer
vectorizer = TfidfVectorizer()
X_vectorized = vectorizer.[1](X)

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

Afit

Btransform

Cfit_transform

Dvectorize

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'transform' without fitting first causes errors.

Using 'fit' alone returns the vectorizer, not transformed data.

To convert raw text into TF-IDF features, use 'fit_transform' which fits the vectorizer and transforms the data in one step.

4fill in blank

hard

Fill both blanks to train a logistic regression model and predict sentiments.

NLP

from sklearn.linear_model import LogisticRegression
model = LogisticRegression()
model.[1](X_vectorized, y)
predictions = model.[2](X_vectorized)

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

Afit

Bpredict

Ctransform

Dscore

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'score' instead of 'predict' to get predictions.

Trying to 'transform' data with the model.

First, train the model using 'fit'. Then, use 'predict' to get sentiment predictions on the data.

5fill in blank

hard

Fill all three blanks to compute accuracy, precision, and recall for the model.

NLP

from sklearn.metrics import [1], [2], [3]
accuracy = accuracy_score(y, predictions)
precision = precision_score(y, predictions, average='macro')
recall = recall_score(y, predictions, average='macro')

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

Aaccuracy_score

Bprecision_score

Crecall_score

Df1_score

Attempts:

3 left

💡 Hint

Common Mistakes

Importing 'f1_score' instead of one of the required metrics.

Forgetting to import all three metrics.

To evaluate the model, import 'accuracy_score', 'precision_score', and 'recall_score' from sklearn.metrics.