Skip to main content
  1. Public practice questions
  2. AB-900 Practice Exams | Copilot and Agent Administration Fundamentals
  3. AB-900 Practice Exam 11

Question 15 — AB-900 Practice Exam 11

An IT admin wants to understand what improves Copilot’s ability to return relevant results without relying on exact keywords. Which three Microsoft Graph mechanisms contribute to meaning-based retrieval and ranking? (Select three answers)

  1. Static keyword-only search across mailboxes

  2. Semantic index built from vectors

  3. Embeddings that represent content meaning

  4. Deep semantic understanding and contextual signals for ranking

Correct answers

  1. Semantic index built from vectors

  2. Embeddings that represent content meaning

  3. Deep semantic understanding and contextual signals for ranking

Explanations

Correct:
Semantic index built from vectors
Graph transforms content into vector representations so meaning-based retrieval can occur even when wording differs.
Embeddings that represent content meaning
Embeddings capture semantic relationships among documents, emails, meetings, and users, enabling intelligent matching.
Deep semantic understanding and contextual signals for ranking
Graph uses semantic understanding plus context such as activity and relationships to refine and rank results.

Wrong:
Static keyword-only search across mailboxes
Graph retrieval is described as semantic and context-aware rather than a simple keyword-only search model.