A machine learning specialist needs to analyze comments on a news website with users across the globe. The specialist must find the most discussed topics in the comments that are in either English…

Question

Options

• AUse an Amazon SageMaker BlazingText algorithm to find the topics independently from
• BUse an Amazon SageMaker seq2seq algorithm to translate from Spanish to English, if necessary.
• CUse Amazon Translate to translate from Spanish to English, if necessary.
• DUse Amazon Translate to translate from Spanish to English, if necessary.
• EUse Amazon Translate to translate from Spanish to English, if necessary.

• B
  84% (31)
• C
  3% (1)
• D
  5% (2)
• E
  8% (3)

Explanation

This 'choose two' question requires two steps: (1) language normalization and (2) topic discovery. Step 1 involves translating Spanish comments to English so all text is in one language before analysis. Option B uses the Amazon SageMaker seq2seq algorithm to perform this translation. The second step involves topic modeling - using an algorithm like Latent Dirichlet Allocation (LDA), also available as a built-in SageMaker algorithm - to discover the most discussed topics across all comments. The seq2seq algorithm in SageMaker is a general encoder-decoder framework suitable for translation tasks when a pre-trained translation model is not desired (vs. Amazon Translate which is the managed translation service). The two-step pattern of translate-then-model-topics is the standard NLP pipeline for multilingual topic discovery. The combination of seq2seq for translation and LDA for topic modeling represents a complete, AWS-native solution.

No community discussion yet for this question.