Question

A travel company receives thousands of customer reviews daily in English, Spanish, French, and Japanese. The product team wants to automatically classify the overall positive or negative tone of each review and identify which entities (hotels, airlines, destinations) are mentioned, with the associated sentiment for each entity. The team has no ML expertise and wants the fastest path to production without training or managing any models. Which Google Cloud approach best meets these requirements?

Accepted Answer

Use the Cloud Natural Language API's entity sentiment analysis and sentiment analysis methods, which support the required languages out of the box.

Answer

Train an AutoML Text classification model in Vertex AI on a labeled sample of reviews for each language.

Answer

Build a BigQuery ML logistic regression model using bag-of-words features extracted from the review text.

Answer

Fine-tune a Gemini model in Model Garden to output sentiment labels and entity mentions for each review.