Using AI to Analyze Reported Defects CT-AI Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   Which AI method is best suited for discovering groups of similar defects without predefined categories?

   • AReinforcement learning
   • BSupervised classification
   • CUnsupervised clustering
   • DLinear regression
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A company's defect management tool stores defect reports as free-text descriptions. The QA lead wants to automatically extract the affected component, the operating system, and the error message from each report to populate structured fields. Which AI technique should be used?

   • AClustering to group similar defect reports together.
   • BTopic modeling to identify the main themes discussed in the defect reports.
   • CSentiment analysis to determine the emotional tone of the defect report.
   • DNamed Entity Recognition (NER) to identify and extract specific entities like component names, OS versions, and error codes.
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   Which task is an example of AI automating defect analysis?

   • AGrouping similar defect reports to identify a common underlying cause
   • BWriting a new test case to reproduce a reported defect
   • CManually assigning a severity level to each new defect report
   • DUpdating the defect tracking tool's database schema
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   What is a key advantage of using AI for defect triage?

   • AIt ensures that all defects are fixed in the order they were reported.
   • BIt can quickly analyze many defects and recommend a prioritization order.
   • CIt removes the need for human testers to review any defects.
   • DIt automatically fixes the highest-priority defects.
   Reveal Solution Discussion Select an answer first
6. 10expert · hard

   A QA team uses an AI clustering tool to group defect reports. After several weeks, they notice that the tool groups defects that are superficially similar (e.g., both mention 'login') but have different root causes (e.g., one is a UI issue, the other is a backend authentication failure). The team wants to improve the clustering to separate these distinct root causes. What is the best approach?

   • AEnrich the clustering input with additional features such as the affected component, stack trace, and test environment.
   • BUse a different similarity metric, such as Euclidean distance instead of cosine similarity.
   • CManually review all clustered defects and correct any misclassifications.
   • DIncrease the number of clusters to force the algorithm to create more groups.
   Reveal Solution Discussion Select an answer first