Question #1869
A financial institution needs to model complex relationships between transactions, accounts, and entities to detect fraudulent activities. The security team must query these relationships efficiently and enforce real-time fraud detection policies with minimal maintenance effort.
Which solution will meet these requirements with the LEAST operational overhead?
Use Amazon RDS to store the data. Use SQL to query relationships and detect fraud.
Use Amazon Neptune to store the data. Use SPARQL to traverse relationships and identify fraudulent patterns.
Use Amazon Redshift to store the data. Use SQL to analyze large datasets for fraud detection.
Use Amazon DynamoDB to store the data. Use PartiQL to query transactional data for anomalies.
Explanation
The correct answer is B. Amazon Neptune is purpose-built for handling complex, interconnected data through graph models, which is critical for fraud detection involving transactions, accounts, and entities. SPARQL allows efficient traversal of relationships to identify patterns, such as fraudulent activities, in real time.
- Option A (RDS/SQL): Relational databases struggle with deep joins and complex relationships, leading to slower queries and higher maintenance for schema optimization.
- Option C (Redshift/SQL): Redshift is designed for analytical workloads, not real-time transactional queries, and lacks native graph traversal capabilities.
- Option D (DynamoDB/PartiQL): DynamoDB excels at key-value access but is inefficient for querying interconnected data, requiring complex data modeling for relationship-heavy use cases.
Neptune's managed service reduces operational overhead while providing the scalability and performance needed for real-time fraud detection. Key takeaway: Use graph databases (Neptune) for complex relationship analysis and real-time traversal.
Answer
The correct answer is: B