MLA-C01 Practice Question 95: Neural Topic Model
AWS Certified Machine Learning Engineer – Associate (MLA-C01) · Test 10 · question 5 of 10
A news publisher has an archive of 2 million articles with no subject labels. The editors want to discover about 50 recurring themes across the archive and give each article soft labels showing how strongly it relates to each theme, since many articles cover several themes at once. The data science team wants to use a SageMaker AI built-in algorithm.
Which algorithm should the team use?
Show answer and explanation
Answer:
SageMaker AI Neural Topic Model (NTM) is an unsupervised algorithm that organizes a corpus of documents into *topics*, groups of words that occur together. The topics are learned from word distributions, so no labels are needed. Documents can be indexed or searched by their soft topic labels, and each document gets a latent representation over the learned topics. That matches an unlabeled archive where articles mix several themes (option B).
- A is wrong because supervised text classification needs labeled training examples, and these articles have no subject labels. - C is wrong because k-means gives each article exactly one cluster. It cannot show that an article is partly about politics and partly about the economy, which the editors explicitly need. - D is wrong because a multiclass classifier also needs labels, and it predicts a single class per article rather than a mix of themes.
Reference: Neural Topic Model (NTM) Algorithm