AI-102 Question 203
Single answerYou are designing an Azure AI solution for a retail company that wants to process customer reviews to extract sentiment, key phrases, and named entities using Azure Cognitive Services. The company also needs to preprocess the incoming data, such as removing special characters, handling missing values, and normalizing text before sending it to the Text Analytics API. Which Azure service or feature should you include in your architecture to handle this preprocessing step effectively?
- A
Azure Data Factory
- B
Azure Synapse Analytics
- C
Azure Cognitive Services - Text Analytics
- D
Azure Functions
Show answer and explanation
Correct answer: D
Explanation
Azure Functions is the best choice for performing custom preprocessing tasks on text data before sending it to the Text Analytics API. While other services like Azure Data Factory or Synapse Analytics are powerful, they are not tailored for the specific requirements of text preprocessing. Azure Cognitive Services - Text Analytics is essential for analyzing text but does not handle preprocessing. Azure Functions provides the flexibility to write and execute custom code for these preprocessing steps efficiently.
- A. Incorrect.
Azure Data Factory is primarily used for data integration and orchestration, and while it can move and transform data, it is not ideal for text-specific preprocessing tasks.
- B. Incorrect.
Azure Synapse Analytics is a powerful analytics service but is not designed specifically for preprocessing text data. It is more suited for large-scale data analysis and data warehousing.
- C. Incorrect.
Azure Cognitive Services - Text Analytics is used for extracting sentiments, key phrases, and named entities, but it does not include built-in functionality to preprocess text before analysis.
- D. Correct.
Azure Functions is a serverless compute option that allows developers to write custom code for specific tasks, such as preprocessing text data (e.g., removing special characters, handling missing values, and normalizing text). This makes it the most suitable option for preprocessing text in this scenario.