引導式測驗 · 中等 等級

AI 產品分類與分類映射測驗

測試 AI 目錄分類:分層路徑、跨分類映射、置信度、特定類別錯誤、分類版本和人工更正。

相關引導路徑AI產品分類
問題 1 的 8

A classifier places a drill under Home Decor instead of Power Tools. Why can this be more than a labeling mistake?

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。

  1. A classifier places a drill under Home Decor instead of Power Tools. Why can this be more than a labeling mistake?
    • A category never changes where shoppers see an item
    • The category can affect navigation, filters, feeds, recommendations, and reporting
    • Only image color depends on category
    • The taxonomy changes the product’s physical features
  2. How does taxonomy mapping differ from classifying within one hierarchy?
    • Mapping always uses images and classification never does
    • Classification translates between languages only
    • The two terms mean exactly the same operation
    • Mapping reconciles labels between schemas, while classification selects a node in a known taxonomy
  3. Why is a hierarchical category error not always equally severe at every level?
    • A wrong top-level branch can make an item much harder to find than a neighboring leaf error
    • All category levels have identical customer effects
    • Leaf categories are not used in navigation
    • A parent category controls only image quality
  4. Low-confidence category predictions should be handled how?
    • Publish the top prediction without review
    • Delete the product from the catalog
    • Route them to a catalog reviewer rather than forcing an uncertain label
    • Increase confidence by repeating the same prediction
  5. What does Google’s googleProductCategory field represent?
    • The product’s customer rating
    • Google’s taxonomy category for the item, distinct from a retailer’s internal product type
    • A machine-generated confidence score
    • The only category a retailer may use
  6. Which metric provides more useful detail than a single overall category score?
    • Accuracy and confusion by category depth and neighboring categories
    • Total model parameter count
    • Number of catalog pages
    • Average product price
  7. Why should taxonomy mappings be versioned?
    • Versioning improves product image resolution
    • Categories never change
    • A model cannot process old products
    • A schema change can alter category meanings and invalidate earlier mappings
  8. What should happen after staff correct a high-impact misclassification?
    • Keep the wrong label for model consistency
    • Delete the source description
    • Update the item and record the correction so the mapping process can be audited and improved
    • Use the correction only for one report