A healthcare company ingests patient data from multiple data sources and stores it in an Amazon S3 staging bucket. An AWS Glue ETL job transforms the data, which is written to an S3-based data lake…

Question

Options

• AUse Amazon Macie pattern matching as part of the ETLjob
• BTrain and use the AWS Glue PySpark filter class in the ETLjob
• CPartition tables and use the ETL job to partition the data on patient name
• DTrain and use the AWS Glue FindMatches ML transform in the ETLjob

• A
  4% (1)
• B
  13% (3)
• C
  8% (2)
• D
  75% (18)

Explanation

The AWS Glue FindMatches ML transform is purpose-built for entity resolution and record linkage - matching records that refer to the same real-world entity even when they lack a common unique identifier, have inconsistent formatting, typos, or missing fields. It is trained on labeled examples of matching and non-matching records and then applied within a Glue ETL job. This is exactly the requirement: matching patient records without a common unique identifier. Amazon Macie (A) is a data security service for detecting sensitive data such as PII, not for matching records. The PySpark filter class (B) filters records based on conditions but cannot perform fuzzy entity matching. Partitioning by patient name (C) assumes names are consistent and identical, which is not reliable for patient record matching across sources.

No community discussion yet for this question.