Prototyping Sentiment Analysis with Agent Platform AutoML
Short answer
B Use Agent Platform AutoML to train a multi-class classification model that predicts sentiment labels based on the training data.
Question
Key Takeaways
The trap is choosing between a generic API (C) or custom deep learning (D); the key insight is recognizing that AutoML (B) balances speed, cost, and accuracy by leveraging existing labeled data without manual coding.
To quickly prototype a sentiment analysis model for ecommerce reviews at minimal cost, use Agent Platform AutoML to train a multi-class classification model on your labeled dataset.
Common mistake: Candidates often choose Option C (Natural Language API) because it is the fastest to implement, but it fails to utilize the company's specific labeled dataset, leading to lower accuracy in this context.
Expert Analysis
The scenario requires a sentiment analysis model to classify reviews into three distinct categories: positive, negative, or neutral. This is a classic multi-class classification problem. The constraints are speed ('quickly prototype') and cost minimization, with a large labeled dataset already available.
Option B is correct because Agent Platform AutoML allows users to train custom models without writing code. It automatically handles feature engineering, model selection, and hyperparameter tuning based on the provided labeled data. Since the company has specific customer review data, training a custom model ensures higher accuracy than generic APIs, while being faster and cheaper than building a deep learning architecture from scratch.
Option A is incorrect because it suggests using a regression model. Regression predicts continuous numerical values (scores), whereas sentiment labels are discrete classes. Furthermore, vectorizing text manually via an Embeddings API adds unnecessary complexity for a prototyping phase where AutoML can handle raw text inputs directly.
Option C is incorrect because the Natural Language API is a pre-trained, generic service. While fast, it does not leverage the company's proprietary, large labeled dataset of customer reviews. Using it would ignore the valuable domain-specific context that could improve accuracy, and the prompt explicitly mentions having a dataset to use.
Option D is incorrect because fine-tuning a BERT-based model requires significant development effort, infrastructure setup, and computational resources. This violates the 'minimize time and cost' constraint compared to the managed AutoML solution.
Exam Strategy
When a question emphasizes 'quickly prototype' and 'minimize cost' while providing a labeled dataset, prioritize managed ML services like AutoML over both generic APIs and custom deep learning frameworks. Generic APIs lack domain specificity, while custom models require excessive time and compute resources. Always match the task type (classification vs. regression) to the algorithm; do not use regression for discrete label prediction.
Frequently Asked Questions
Why not use the Natural Language API for sentiment analysis?
Can I use regression for sentiment analysis?
Community Votes
100% of anonymous learners picked answer B. Votes are pick records left by other test-takers — they are not the verified answer.
Discussion 2 contributions
Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.
From public community threads 2
-
taylz876 👍 1 Answered BSimilar to Q68 and Q12 Why B is right • You already have a large labeled dataset (positive, negative, neutral), so a model trained on your own reviews will be accurate on your kind of text. • Quick prototype with low cost: AutoML text classification needs no code, and it handles the model and tuning for you. Why the others are wrong • A: Embeddings plus a regression model predicts a score, not the three labels, and it's more steps. • C: The Natural Language API gives general sentiment scores without using your labels. It's quick, but it isn't tuned to your reviews, so it's less accurate for "accurately predicts the sentiment labels". • D: Fine-tuning BERT means writing and managing code, which costs more time.
-
n2183712847 👍 1 Answered BCorrect answer: B. Agent Platform AutoML can directly train a multi-class classification model using the labeled reviews, providing a fast, managed, and cost-efficient way to prototype sentiment prediction; A incorrectly treats the task as regression, C uses a generic prebuilt sentiment API rather than leveraging the company’s labeled dataset, and D requires significantly more development and training effort than AutoML.
Reader comments & corrections
No reader corrections yet — spotted an error or have a note? Share it below.
Related Analysis
Practice All PMLE Questions
Access 65 questions with complete answers and detailed explanations.
View Full PMLE Practice Test →