引導式測驗 · 中等 等級
用於圖書館編目和元資料測驗的人工智慧
了解人工智慧如何建議圖書館元資料、為什麼編目標準和權限控制很重要,以及工作人員如何防止偏見或不準確的記錄。
問題 1 的 8
What role can AI play in cataloging?
預覽每個問題8 問題
當您參加測驗時,就會出現答案和解釋。
- What role can AI play in cataloging?
- Suggest fields or terms for cataloger review
- Replace cataloging standards
- Determine an author’s identity without evidence
- Guarantee that a record is culturally accurate
- Why verify an AI-suggested MARC field?
- MARC fields have no defined structure
- Any field is valid if the title is correct
- A plausible field may still use an incorrect tag or subfield
- The model automatically checks all indicators
- What can authority control help prevent?
- Confusing a new edition with a new creator
- Confusing distinct creators or names
- Every biased description
- The need for cataloger review
- Why review machine-generated subject terms?
- Subject terms are independent of item content
- Generated terms are always more current
- Review makes metadata harder to search
- They can reflect outdated language or misrepresent the work
- Why is an incorrect entity merge serious?
- It can collapse distinct people, editions, or works
- It changes display formatting but cannot affect attribution or discovery
- It improves authority control automatically
- It affects only model training
- What should bulk metadata changes include?
- A guarantee that AI never makes mistakes
- Only the final number of records
- Sampling and a way to roll back errors
- No review because the change is large
- Which provenance detail supports later review?
- Only the catalog user’s screen size
- Which fields were generated and how staff changed them
- The model’s marketing name
- The number of records in the database
- How should libraries handle AI-suggested summaries?
- Publish them automatically if grammatically fluent
- Assume the model read the complete work
- Treat them as a substitute for subject analysis
- Compare them with the item and correct unsupported claims