DEA-C01 Question 277
Select 3You are designing a data lake architecture for an e-commerce company that collects transactional data, product catalog information, and customer reviews. The transactional data is structured, the product catalog is semi-structured (stored as JSON files), and customer reviews contain unstructured text data. How should you model and store these datasets in your data lake to enable efficient querying and analytics?
- A
Store transactional data in a columnar format such as Parquet or ORC for optimized analytics.
- B
Store the product catalog JSON files as-is in an Amazon S3 bucket and use Amazon Athena to query the data.
- C
Convert customer reviews to a structured format and store them as CSV files for easier querying.
- D
Store customer reviews as raw text files in Amazon S3 and use Amazon Comprehend for text analysis.
- E
Use Amazon DynamoDB to store all datasets to ensure consistent performance across structured, semi-structured, and unstructured data.
Show answer and explanation
Correct answers: A, B, D
Explanation
When designing a data lake, it is crucial to select the appropriate storage and processing methods for each data type. Structured data benefits from columnar storage formats for optimized analytics, semi-structured data can be stored as-is and queried with schema-on-read services like Amazon Athena, and unstructured data should be stored in its raw form for flexibility in analysis. Using specialized AWS services such as Amazon Comprehend for unstructured data processing ensures the architecture is efficient and scalable.
- A. Correct.
This is correct because structured data like transactional records benefits from a columnar storage format such as Parquet or ORC, which are optimized for analytical queries and reduce storage costs.
- B. Correct.
This is correct because semi-structured data like JSON can be stored as-is in Amazon S3, and services like Amazon Athena allow schema-on-read to query the data efficiently without transformation.
- C. Incorrect.
This is incorrect because unstructured data like customer reviews is not naturally suited for conversion to a structured format like CSV. Converting unstructured text to CSV can result in a loss of valuable context and flexibility.
- D. Correct.
This is correct because unstructured text data, such as customer reviews, can be stored as raw text in Amazon S3 and processed using AWS services like Amazon Comprehend for natural language processing and sentiment analysis.
- E. Incorrect.
This is incorrect because Amazon DynamoDB is a NoSQL database designed for low-latency key-value and document store use cases. It is not suitable for all types of data, especially for large-scale analytics on structured and unstructured data.