Prepare and test your skills
Prepare and test your skills
Free sample question, answered and explained below. The Professional Cloud DevOps Engineer practice bank has 1,455 more like it. Try 10 free: instant score, no account.
An organization is streaming application logs from a clinical healthcare microservice into Google Cloud. During a security audit, you identify that unstructured log messages contain both direct identifiers—specifically proprietary Medical Record Numbers (MRNs) following the pattern ###-#-#####—and potential quasi-identifiers such as person names that could expose Protected Health Information (PHI).
You need to design a Sensitive Data Protection (Cloud DLP) inspection strategy that accurately identifies and classifies these distinct data categories while minimizing false positives from standard application terms.
Which inspection configuration should you implement?
BigQuery ML allows users to build, evaluate, and execute machine learning models (such as binary logistic regression) using standard SQL queries inside BigQuery datasets. It is primarily used for predictive data analytics, such as demographic forecasting or tabular classification.
PERSON_NAME).You were performing retrospective analytical batch evaluations on historical tabular log data to predict high-level event classifications rather than configuring real-time sensitive data log inspection rules.
A regular custom dictionary scans for a small, predefined list of static terms in-line. Assigning a fixed VERY_LIKELY likelihood to custom pattern matches forces the detector to treat every pattern occurrence as an absolute match regardless of surrounding context.
PERSON_NAME to a static dictionary means any patient whose name is not pre-populated in the dictionary will be completely missed, severely undermining PHI classification.VERY_LIKELY on generic numeric patterns without evaluating proximity words or checksum validation can lead to high false-positive rates if common hyphenated numbers share similar structures.You were scanning exclusively for a fixed, closed roster of staff members and the numeric format was guaranteed to be globally unique with zero possibility of collisions in log outputs.
A stored custom dictionary detector is designed to match very large lists of discrete words or phrases (up to tens of millions of entries) sourced from Cloud Storage or BigQuery tables, leveraging generated Bloom filters for rapid lookup.
###-#-#####).PERSON_NAME occurrences without context degrades the detector's ability to catch actual PHI patient names in the clinical logs.The requirement was to match fixed, known lists of hundreds of thousands of specific patient full names or clinic facility titles stored in a database table rather than pattern-based MRNs.
Sensitive Data Protection (Cloud DLP) provides inspection mechanisms to locate, classify, and de-identify sensitive data in text streams. While Google Cloud supplies built-in infoTypes (such as PERSON_NAME), organizations often require specialized detection for proprietary formats—such as custom Medical Record Numbers (MRNs)—and nuanced context-aware detection rules to distinguish sensitive quasi-identifiers from general text.
[1-9]{3}-[1-9]{1}-[1-9]{5}, Sensitive Data Protection can identify organizational MRNs that do not conform to standardized national identifiers.PERSON_NAME detector with hotword rules allows the engine to evaluate proximity words such as patient or doctor. Finding these hotwords within a defined search window shifts the match likelihood (e.g., from POSSIBLE to VERY_LIKELY), differentiating true PHI from generic names.REDACTED or placeholder identifiers) prevents benign application tokens from generating erroneous findings.InspectConfig and InspectionRuleSet).This approach aligns with Google Cloud best practices for sensitive data discovery in unstructured logs by combining custom pattern detection for strict direct identifiers with context-sensitive inspection rule sets for broad quasi-identifiers.
Answer them yourself and get your score the moment you finish. No account needed.
Full access includes
Keep the momentum going with these hand-picked practice scenarios
Last updated Validated