Google ADP 試験概要:

Google ADP 試験シラバストピック:

Google Associate Data Practitioner 認定 ADP 試験問題：

問題 #1

Your retail company wants to analyze customer reviews to understand sentiment and identify areas for improvement. Your company has a large dataset of customer feedback text stored in BigQuery that includes diverse language patterns, emojis, and slang. You want to build a solution to classify customer sentiment from the feedback text. What should you do?

A. Develop a custom sentiment analysis model using TensorFlow. Deploy it on a Compute Engine instance.
B. Export the raw data from BigQuery. Use AutoML Natural Language to train a custom sentiment analysis model.
C. Preprocess the text data in BigQuery using SQL functions. Export the processed data to AutoML Natural Language for model training and deployment.
D. Use Dataproc to create a Spark cluster, perform text preprocessing using Spark NLP, and build a sentiment analysis model with Spark MLlib.

問題 #2

You used BigQuery ML to build a customer purchase propensity model six months ago. You want to compare the current serving data with the historical serving data to determine whether you need to retrain the model.
What should you do?

A. Evaluate the data skewness.
B. Compare the confusion matrix.
C. Compare the two different models.
D. Evaluate data drift.

問題 #3

Your organization has decided to move their on-premises Apache Spark-based workload to Google Cloud.
You want to be able to manage the code without needing to provision and manage your own cluster. What should you do?

A. Migrate the Spark jobs to Dataproc Serverless.
B. Configure a Google Kubernetes Engine cluster with Spark operators, and deploy the Spark jobs.
C. Migrate the Spark jobs to Dataproc on Google Kubernetes Engine.
D. Migrate the Spark jobs to Dataproc on Compute Engine.

問題 #4

Your organization's ecommerce website collects user activity logs using a Pub/Sub topic. Your organization's leadership team wants a dashboard that contains aggregated user engagement metrics. You need to create a solution that transforms the user activity logs into aggregated metrics, while ensuring that the raw data can be easily queried. What should you do?

A. Create a Dataflow subscription to the Pub/Sub topic, and transform the activity logs. Load the transformed data into a BigQuery table for reporting.
B. Create a Cloud Storage subscription to the Pub/Sub topic. Load the activity logs into a bucket using the Avro file format. Use Dataflow to transform the data, and load it into a BigQuery table for reporting.
C. Create an event-driven Cloud Run function to trigger a data transformation pipeline to run. Load the transformed activity logs into a BigQuery table for reporting.
D. Create a BigQuery subscription to the Pub/Sub topic, and load the activity logs into the table. Create a materialized view in BigQuery using SQL to transform the data for reporting

問題 #5

Your organization has a BigQuery dataset that contains sensitive employee information such as salaries and performance reviews. The payroll specialist in the HR department needs to have continuous access to aggregated performance data, but they do not need continuous access to other sensitive dat a. You need to grant the payroll specialist access to the performance data without granting them access to the entire dataset using the simplest and most secure approach. What should you do?

A. Create a table with the aggregated performance data. Use table-level permissions to grant access to the payroll specialist.
B. Create a SQL query with the aggregated performance data. Export the results to an Avro file in a Cloud Storage bucket. Share the bucket with the payroll specialist.
C. Create row-level and column-level permissions and policies on the table that contains performance data in the dataset. Provide the payroll specialist with the appropriate permission set.
D. Use authorized views to share query results with the payroll specialist.

解説：